From andrew at ceruleansystems.com Mon Oct 19 00:58:44 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Mon, 19 Oct 2009 00:58:44 -0700 Subject: [FoRK] Adverse selection in insurance In-Reply-To: <4ADBFDE5.4030901@lig.net> References: <830B3CB3-40EF-421C-8590-2DD3471BC4DC@place.org> <4ADB5E4A.30804@b3k.us> <4ADB6F8A.6020709@boxbe.com> <4ADB70F4.2070202@b3k.us> <4ADBE1D9.2050708@lig.net> <5F02C536-3138-499B-8043-86528D6180F4@ceruleansystems.com> <4ADBFDE5.4030901@lig.net> Message-ID: <9EFCE066-C93A-4DAB-832B-585695D0E9B0@ceruleansystems.com> On Oct 18, 2009, at 10:49 PM, Stephen Williams wrote: > Everyone who complains about dysfunctional insurance is to blame for > it? In the sense that they applaud or ignore the very actions that create the dysfunction. Ignorance of basic economics is no defense. > What regulations encourage it? I find that it is more a natural > consequence of insurance companies having to negotiate with most of > their clients (large companies / agencies) and of individuals having > no way to negotiate effectively. While regulations twist many > aspects of the situation, that specific problem seems more caused by > a lack of regulations. You might argue that the heavy-handed regulations that currently exist are stupid, but they most certainly do exist in vast quantity. Furthermore, they essentially mandate that "insurance" is extraordinarily wasteful to no obvious end beyond benefiting various special interests; large companies can absorb that kind of waste, small companies and individuals cannot. I always marvel that the regulatory agencies that literally define the market are never held to be responsible for the markets they define. I'm sorry, you are going to have to accept the fact that the industry is heavily regulated at multiple levels and that the regulation has been an abject failure, very arguably worse than minimal regulation. >> I'm sorry, I thought we were talking about insurance. My mistake. > > I was talking about insurance for high-risk individuals. Those that > are already costing a lot are in an different category since they > are no longer a risk, they are a reality. High-risk individuals have a straightforward mathematical cost, that is promptly ignored by insurance regulations. It is akin to legislating the value of pi. People do not like the mathematical reality, so they pretend they can legislate a new one. As long as governments declare that fiction is the new reality, we will continue to have adverse outcomes that seem inexplicable to the innumerate masses. This kind of nonsense happens with all insurance, but it is particularly bad with medical insurance. A compelling argument can be made that it would be far cheaper to have literal, true insurance without the morass of venal and idiotic regulations and then pick up the remainder with straightforward welfare. Unfortunately, the government has decided to redefine insurance to be welfare with a dollop of heavy special-interest largesse thrown on top. See: current "health care reform" in Congress. > If we were talking about auto or house insurance, perhaps losses > beforehand are just too bad. (Of course, we often invent aid > methods even then in large cases.) With live people, it just > doesn't apply the same way. Nonsense, we apply this to houses the exact same way we apply it to medical care. The government subsidizes the insurance of people who live in high-risk areas who don't want to pay high-risk premiums without any regard for whether or not they can afford it. There is no moral hazard so large that the population will not vote for it if they think they can extract free stuff from their fellow citizens. It never saves any net money. > Is health care a right? To what degree and in what sense? Health care can't be a right. Health care is a "right" insofar as the community feels like delivering it. It is not as though you are guaranteed to get the best health care possible, and in practice it is more like a minimum level of service with little investment in improving long-term outcomes (e.g. biomedical R&D). If you get seriously ill in a country with medicine that asserts this "right", it is not as though you get teleported to Stanford or Mayo for your "right" to a significantly better medical outcome. As nominal rights go, it is a pretty weak fiction in that there is no possible way to deliver it even in principle in most countries. Rights like freedom of expression are fully functional anywhere on the globe without any investment by other people. > Do we want true mutual insurance? How is it limited? Do we allow > people to opt out? Do we penalize them now or later? If it is actually insurance, it doesn't cost you anything if people opt out. If it costs money if some class of people opt out, it is a sure sign that you've allowed the ignorati to design your regulations or at the very least incorporated significant deception. > Do we penalize people for lifestyle? (Smoking, what they eat, > weight, too much alcohol, drugs, etc.) How? Why not, we micromanage every other aspect of people's lives for lesser reasons. You have no right to these choices. > How do we keep funding effective medical research in a lucrative > way, create true competition for appropriate goals such as ever more > efficient and effective care, while having fair mutual insurance > with little overhead? Insurance can be fixed by letting it be insurance. In essence, stop mandating that I have to buy insurance that *must* cover 27 flavors of cultural anti-science woo that I can consume on a monthly basis if I so choose. FFS, the insurance companies have to insure things that it makes no economic sense to insure pretty much by definition. Nor do I see a reason to be required to pay for all sorts of anti-science nonsense as standard coverage items. At the same time, people have to realize that the proliferation of medical technology is increasing the number of things we can spend medical dollars on, hence why even countries with socialized medicine are seeing costs grow at similar rates to the US. Medical care used to be cheap because medical care used to suck. It is very easy to arrive at the point where it is not possible to supply all possible medical care to everyone. Biomedical R&D is the elephant in the room being vigorously ignored in the current debate. The US does almost all of it, and it is largely done by private companies. The stagnation of biomedical R&D investment would have an extraordinary cost in terms of unnecessary loss of life. Most industrialized countries have abdicated their supposed social responsibility toward the health of their people; it is highly ironic that the baser ideals of the for-profit US system is producing the vast majority of the advances that benefit the average health of humanity. For all the talk about really caring, no one else seems to find the money worth investing. From andrew at ceruleansystems.com Mon Oct 19 01:07:35 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Mon, 19 Oct 2009 01:07:35 -0700 Subject: [FoRK] Adverse selection in insurance In-Reply-To: <4ADC010E.2080908@b3k.us> References: <830B3CB3-40EF-421C-8590-2DD3471BC4DC@place.org> <4ADB5E4A.30804@b3k.us> <4ADB6F8A.6020709@boxbe.com> <4ADB70F4.2070202@b3k.us> <4ADBE1D9.2050708@lig.net> <5F02C536-3138-499B-8043-86528D6180F4@ceruleansystems.com> <4ADC010E.2080908@b3k.us> Message-ID: On Oct 18, 2009, at 11:02 PM, Benjamin Black wrote: > Yes, pity, the poor, put upon insurance companies who can't afford to > spend millions on lobbying legislatures to get exactly the regulations > they want passed, often having written them for the legislators. Congratulations, you've discovered regulatory capture. The real trick is understanding it and learning how to avoid it, but I'm not sanguine. > Instead, the people of the states have risen up to smite them with > their > awesome democratic power. Give it a rest. The notion that insurance > companies, working hand in glove with the health care providers and > big > pharma, are victims rather than perpetrators is ludicrous. The people used their "awesome democratic power" to make it possible, practically begged for it. The special interests were all too happy to oblige. I'm hoping that the people will learn this elementary lesson. Some day. Eventually. > The companies in these industries do not want reform because the > system > we have works extremely well for them. Those rent-seeking special > interests? They're the health care industry. And the people voted to allow this. Clearly a trustworthy lot when it comes to important decisions involving the design of economic systems. From beberg at mithral.com Mon Oct 19 01:10:06 2009 From: beberg at mithral.com (Adam L Beberg) Date: Mon, 19 Oct 2009 01:10:06 -0700 Subject: [FoRK] Adverse selection in insurance In-Reply-To: <4ADC010E.2080908@b3k.us> References: <830B3CB3-40EF-421C-8590-2DD3471BC4DC@place.org> <4ADB5E4A.30804@b3k.us> <4ADB6F8A.6020709@boxbe.com> <4ADB70F4.2070202@b3k.us> <4ADBE1D9.2050708@lig.net> <5F02C536-3138-499B-8043-86528D6180F4@ceruleansystems.com> <4ADC010E.2080908@b3k.us> Message-ID: <4ADC1EDE.1040203@mithral.com> The "death spiral" is how we lost health insurance for dependents at Stanford. It was going to be over 800/person/month, so it went away. Kids are now mostly on the state's Healthy Families system (a trick learned from Walmart) - which almost went away as well. Soon Obama's plan, on orders from the health insurance companies, will mean grad students will be forced to pay for it, and it will get MUCH more expensive. And with California now a failed state... Great for the healthy insurance executives. In the old days it would be good for their shareholders too, but the system is better tuned now. -- Adam L. Beberg http://www.mithral.com/~beberg/ From dave.long at bluewin.ch Mon Oct 19 05:13:59 2009 From: dave.long at bluewin.ch (Dave Long) Date: Mon, 19 Oct 2009 14:13:59 +0200 Subject: [FoRK] Adverse selection in insurance In-Reply-To: <4ADC010E.2080908@b3k.us> References: <830B3CB3-40EF-421C-8590-2DD3471BC4DC@place.org> <4ADB5E4A.30804@b3k.us> <4ADB6F8A.6020709@boxbe.com> <4ADB70F4.2070202@b3k.us> <4ADBE1D9.2050708@lig.net> <5F02C536-3138-499B-8043-86528D6180F4@ceruleansystems.com> <4ADC010E.2080908@b3k.us> Message-ID: <07F1426F-5B01-4822-AA13-04F84541E7A7@bluewin.ch> Just a note on how we do it here in CH: 1/ everyone's covered. why not simply have insurance act as insurance? 2/ there's a gamut of providers and a wide range of plans, but the relationship is directly between the insurer and the insuree. what motivates this US nonsense about tying insurance to (full time?) employment? 3/ any social subsidization is done directly between the insuree and the cantons, in the form of tax credits for premiums paid. In general I find the swiss approach of first setting market prices and then ensuring that no one gets given a raw deal to be much less schizophrenic than the american approach of first claiming universal rights and then quietly delivering very uneven levels of service... -Dave From geege4 at gmail.com Mon Oct 19 08:05:29 2009 From: geege4 at gmail.com (geege schuman) Date: Mon, 19 Oct 2009 11:05:29 -0400 Subject: [FoRK] Ad Age Message-ID: <493a95a00910190805ob325c9dwa2bbba663bdee47a@mail.gmail.com> http://adage.com/mediaworks/article?article_id=139760 Thank you, social media. (Good article - good comments - angst-ridden poem (is there really any other kind?)) Geege From ken_ganshirt at yahoo.ca Mon Oct 19 08:17:59 2009 From: ken_ganshirt at yahoo.ca (Ken Ganshirt @ Yahoo) Date: Mon, 19 Oct 2009 08:17:59 -0700 (PDT) Subject: [FoRK] Adverse selection in insurance In-Reply-To: <5F02C536-3138-499B-8043-86528D6180F4@ceruleansystems.com> Message-ID: <152793.11093.qm@web33002.mail.mud.yahoo.com> --- On Sun, 10/18/09, J. Andrew Rogers wrote: > > That is because regulations frequently do not let insurance > companies sell, you know, insurance.? Part of the > problem is that 95+% of all Americans are deeply ignorant > about the concept of insurance, > ...[snip]... > > The argument has not been about insurance at all, even the > GOP proposals are only marginally more about fixing > insurance.? The argument is over how many Americans are > we going to officially add to the welfare rolls, but people > never like the sound of that so the talking heads blather > incoherent things about "insurance". > Thanks for that perspective. I'm not sure it's because 95+% are ignorant of what insurance really is. I suspect a lot of us are like me: never really thought about it. That's what everyone calls it so you just sort of assume that's what it is. Got bigger alligators biting at the haunches; no need to analyse whether it's really what it's called, you know. Perhaps only 94% are genuinely ignorant. :-) I'm Canadian so I have no problem with socialized single-payor health care. But I can see how inefficient and ineffective it is to try to use regulation to turn one part of the insurance industry into a delivery system for socialized health care rather than simply bite the bullet and do it properly ... assuming that's where one wants to end up, of course. ...ken... __________________________________________________________________ Looking for the perfect gift? Give the gift of Flickr! http://www.flickr.com/gift/ From sdw at lig.net Mon Oct 19 08:27:45 2009 From: sdw at lig.net (Stephen Williams) Date: Mon, 19 Oct 2009 08:27:45 -0700 Subject: [FoRK] Ad Age In-Reply-To: <493a95a00910190805ob325c9dwa2bbba663bdee47a@mail.gmail.com> References: <493a95a00910190805ob325c9dwa2bbba663bdee47a@mail.gmail.com> Message-ID: <4ADC8571.2020500@lig.net> geege schuman wrote: > http://adage.com/mediaworks/article?article_id=139760 > > third-fastest-growing economic sector over the next five years, after mining > and construction. Almost none of that growth is forecast to come from > shrinking traditional media.> > > Thank you, social media. (Good article - good comments - angst-ridden poem > (is there really any other kind?)) > > Geege > See "The Answer Factory" in the current Wired 17.11, not yet online... "A fiendishly clever startup knows what we are Googling - then churns out thousands of cheap videos and articles to meet our every whim and wish. Why the future of online content is fast, disposable, and profitable as hell." All about Demand Media. http://www.demandmedia.com/ "This year, the privately held Demand is expected to bring in about $200 million in revenue; its most recent round of financing by blue-chip investors valued the company at $1 billion." They follow a guy who gets paid about $20/video and has filmed over 40,000 videos for them already. Very interesting and almost scary. But potentially cool too. Stephen From hkpang at gmail.com Mon Oct 19 08:49:54 2009 From: hkpang at gmail.com (HK Pang) Date: Mon, 19 Oct 2009 11:49:54 -0400 Subject: [FoRK] Ad Age In-Reply-To: <493a95a00910190805ob325c9dwa2bbba663bdee47a@mail.gmail.com> References: <493a95a00910190805ob325c9dwa2bbba663bdee47a@mail.gmail.com> Message-ID: <5234e09e0910190849u53ac6b94xaa2fa976b6bb9333@mail.gmail.com> Sort of on a related topic.. does anyone know if any media company is entitled to use derived data to determine the user's undisclosed information? While on facebook, my wife was prompt with this question "description of your activity is confusing, should it be a) xxx modified his profile or b) xxx modified her profile". Obviously, if she answered b, advertisers can determine her gender even through she specifically chose not to disclose it in her facebook profile. So, can facebook legally pass this piece of info to advertisers? On Mon, Oct 19, 2009 at 11:05 AM, geege schuman wrote: > http://adage.com/mediaworks/article?article_id=139760 > > third-fastest-growing economic sector over the next five years, after mining > and construction. Almost none of that growth is forecast to come from > shrinking traditional media.> > > Thank you, social media. ?(Good article - good comments - angst-ridden poem > (is there really any other kind?)) > > Geege > _______________________________________________ > FoRK mailing list > http://xent.com/mailman/listinfo/fork > From ken_ganshirt at yahoo.ca Mon Oct 19 08:58:52 2009 From: ken_ganshirt at yahoo.ca (Ken Ganshirt @ Yahoo) Date: Mon, 19 Oct 2009 08:58:52 -0700 (PDT) Subject: [FoRK] Adverse selection in insurance In-Reply-To: <9EFCE066-C93A-4DAB-832B-585695D0E9B0@ceruleansystems.com> Message-ID: <85390.14991.qm@web33001.mail.mud.yahoo.com> --- On Mon, 10/19/09, J. Andrew Rogers wrote: > > ...[snip]... > > I'm sorry, you are going to have to accept the fact that > the industry is heavily regulated at multiple levels and > that the regulation has been an abject failure, very > arguably worse than minimal regulation. > I'm not arguing or debating, just interested. "...the regulation has been an abject failure" at what, exactly? Has anyone defined, or analyzed after the fact, what all that regulation of health "insurance" is supposed to achieve? To this tiny reductive mind, looking in from the outside, it looks like someone is trying to turn the health part of the US insurance industry into socialized medicine. If that is the goal, the regulation seems to be working to some extent, yes? Or is there some other goal(s) for all that regulation? > > > How do we keep funding effective medical research in a > lucrative way, create true competition for appropriate goals > such as ever more efficient and effective care, while having > fair mutual insurance with little overhead? > > ...[snip]... > > Biomedical R&D is the elephant in the room being > vigorously ignored in the current debate.?... > I don't understand. The current debate in this thread seems to be about whether American health insurance is really insurance (it's looking more and more like, probably not). The current debate in the broader sense, in America, seems to be: - whether socialized single-payor medicine is a Good Thing or not. - whether tweaks to health care insurance regulations is the way to deliver [more] socialized health care, versus some other method, if you agree that you want to head down that road. But what does the method of paying for health care have to do with biomedical R&D (potential for stagnation, you said)? Regardless how American health care gets paid for, won't the biomedical R&D companies still have a market for all the stuff they invent? ...ken... __________________________________________________________________ Yahoo! Canada Toolbar: Search from anywhere on the web, and bookmark your favourite sites. Download it now http://ca.toolbar.yahoo.com. From andrew at ceruleansystems.com Mon Oct 19 09:00:25 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Mon, 19 Oct 2009 09:00:25 -0700 Subject: [FoRK] Adverse selection in insurance In-Reply-To: <07F1426F-5B01-4822-AA13-04F84541E7A7@bluewin.ch> References: <830B3CB3-40EF-421C-8590-2DD3471BC4DC@place.org> <4ADB5E4A.30804@b3k.us> <4ADB6F8A.6020709@boxbe.com> <4ADB70F4.2070202@b3k.us> <4ADBE1D9.2050708@lig.net> <5F02C536-3138-499B-8043-86528D6180F4@ceruleansystems.com> <4ADC010E.2080908@b3k.us> <07F1426F-5B01-4822-AA13-04F84541E7A7@bluewin.ch> Message-ID: On Oct 19, 2009, at 5:13 AM, Dave Long wrote: > Just a note on how we do it here in CH: > > 1/ everyone's covered. why not simply have insurance act as > insurance? > > 2/ there's a gamut of providers and a wide range of plans, but the > relationship is directly between the insurer and the insuree. what > motivates this US nonsense about tying insurance to (full time?) > employment? > > 3/ any social subsidization is done directly between the insuree and > the cantons, in the form of tax credits for premiums paid. > > In general I find the swiss approach of first setting market prices > and then ensuring that no one gets given a raw deal to be much less > schizophrenic than the american approach of first claiming universal > rights and then quietly delivering very uneven levels of service... Yeah, as such things go the Swiss system is tidy and well-designed. There is a significant path dependency component to the implementation of systems in various countries. The US ended up with employer-based health insurance -- a bad idea -- as a direct market consequence of ill-conceived regulation under Roosevelt and has never been able to get away from it since even though the regulations that caused it have not existed in a very long time. This kind of path dependency is one of the bigger reasons not to cluelessly monkey with heavy-handed regulation. Even if you undo it later, you may find the market on an unfortunate trajectory that is politically very difficult to correct. It is not as though being wrong does not have a long-term cost, but something resembling critical analysis of long-term consequences is rarely done. From andrew at ceruleansystems.com Mon Oct 19 09:38:41 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Mon, 19 Oct 2009 09:38:41 -0700 Subject: [FoRK] Adverse selection in insurance In-Reply-To: <85390.14991.qm@web33001.mail.mud.yahoo.com> References: <85390.14991.qm@web33001.mail.mud.yahoo.com> Message-ID: On Oct 19, 2009, at 8:58 AM, Ken Ganshirt @ Yahoo wrote: > --- On Mon, 10/19/09, J. Andrew Rogers > wrote: >> >> ...[snip]... >> >> I'm sorry, you are going to have to accept the fact that >> the industry is heavily regulated at multiple levels and >> that the regulation has been an abject failure, very >> arguably worse than minimal regulation. >> > > I'm not arguing or debating, just interested. "...the regulation has > been an abject failure" at what, exactly? The nominal purpose of regulation is to create well-functioning markets. This usually requires a relatively minimal set of intelligent regulations. Instead, we have hyper-regulated markets that are mediocre at best. In practice, regulation is mostly about rent-seeking and political control. If we end up with a reasonably well-functioning market, that is just a bonus. > To this tiny reductive mind, looking in from the outside, it looks > like someone is trying to turn the health part of the US insurance > industry into socialized medicine. If that is the goal, the > regulation seems to be working to some extent, yes? Heh, I suppose you could frame it that way. >> >> Biomedical R&D is the elephant in the room being >> vigorously ignored in the current debate. ... >> > > I don't understand. The current debate in this thread seems to be > about whether American health insurance is really insurance (it's > looking more and more like, probably not). As a side-effect of the US market, most of the world's biomedical R&D happens in the US. New medical treatments, devices, and drugs are developed and tested in the US first because there is a market where they are allowed to make a nice profit (about 10-15% profit for the biomedical tech industry). There is something disconcerting about the fact that biomedical R&D has all but disappeared in the rest of the industrialized world. Since the biomedical R&D in the US is primarily funded and driven by profit motive, it generally follows that significantly reducing the average profits from biomedical R&D in the US will greatly reduce activity in that sector. R&D dollars will go toward more profitable technology sectors. No one can make the argument that socialized medicine coexists with continued strong biomedical R&D investment because the empirical evidence very strongly points in the opposite direction. > But what does the method of paying for health care have to do with > biomedical R&D (potential for stagnation, you said)? US companies heavily invest in biomedical R&D under the notion that can leverage that R&D for a profit. If the government engages in price-fixing to reduce its own costs, that greatly reduces the probability of a profit from high-risk R&D. The US government already improperly uses the FDA to get "free" stuff from the biomedical technology companies, and there is no reason to believe this type of behavior will not continue if they have more control of the system. > Regardless how American health care gets paid for, won't the > biomedical R&D companies still have a market for all the stuff they > invent? A market yes, but not a worthwhile profit. In most industrialized countries, the allowed profit margins do not justify continued biomedical R&D. If all industrialized countries have similar profit margins, biomedical R&D will wither because the aggregate return on investment will plummet. From gojomo at boxbe.com Mon Oct 19 09:50:56 2009 From: gojomo at boxbe.com (Gordon Mohr) Date: Mon, 19 Oct 2009 09:50:56 -0700 Subject: [FoRK] Adverse selection in insurance In-Reply-To: References: <830B3CB3-40EF-421C-8590-2DD3471BC4DC@place.org> <4ADB5E4A.30804@b3k.us> <4ADB6F8A.6020709@boxbe.com> <4ADB70F4.2070202@b3k.us> <4ADBE1D9.2050708@lig.net> <5F02C536-3138-499B-8043-86528D6180F4@ceruleansystems.com> <4ADC010E.2080908@b3k.us> <07F1426F-5B01-4822-AA13-04F84541E7A7@bluewin.ch> Message-ID: <4ADC98F0.6010502@boxbe.com> J. Andrew Rogers wrote: > There is a significant path dependency component to the implementation > of systems in various countries. The US ended up with employer-based > health insurance -- a bad idea -- as a direct market consequence of > ill-conceived regulation under Roosevelt and has never been able to get > away from it since even though the regulations that caused it have not > existed in a very long time. All true, but it is also important to note: One of the biggest distortions that bias the US towards an emplyer-provided system still exists: employer-paid health insurance costs are free of income taxes; individual-paid insurance costs are not. We don't even know what a US individual-insurance market would look like because it's been placed at a ~15-40% cost disadvantage to an employer-centric system by tax policy. An incremental approach to fixing the problems in US health coverage, one by one, and letting people and companies adjust should start with removing that tax policy distortion. A few years later, we might find a vibrant competitive market for portable health insurance, making the remaining problems easier to address. Or not. But until it's tried we don't know. As far as I can tell, there are a couple big reasons the ruling party doesn't consider this first step: (1) the ruling party's key union constituency has negotiated a lot of high-end health coverage dependent on unlimited deductibility of employer-paid health insurance -- and extending this same unlimited deductibility to individual insurance would be a budget-buster. (Capping the deductibility could make equal treatment of employer- and individual- costs revenue-neutral, but such 'progressive' taxation of health benefit income is resisted by the unions who are 'rich' in this one dimension.) (2) the ruling party prefers the system to get worse before it gets better, to feed a useful crisis mentality, helping them finally get the single-payer system they really want - Gordon From andrew at ceruleansystems.com Mon Oct 19 09:54:29 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Mon, 19 Oct 2009 09:54:29 -0700 Subject: [FoRK] Adverse selection in insurance In-Reply-To: References: <85390.14991.qm@web33001.mail.mud.yahoo.com> Message-ID: <8E798C62-3718-416C-BDCB-02AC66BCF623@ceruleansystems.com> On Oct 19, 2009, at 9:38 AM, J. Andrew Rogers wrote: > > No one can make the argument that socialized medicine coexists with > continued strong biomedical R&D investment because the empirical > evidence very strongly points in the opposite direction. I should add that even to the extent there is government R&D funding, which is a pretty small fraction of the total pie, that government R&D funding seems to be motivated at least in part by the commercial lobby. As in, government biomedical R&D funding exists primarily to deliver money to important constituents, not to do biomedical R&D for its own sake. If we eliminate or reduce the well-connected constituents pushing for government R&D funding, it follows that this will no longer be a funding priority. This probably has a lot to do with why EU biomedical R&D funding by governments is paltry compared to US government biomedical R&D funding. From geege4 at gmail.com Mon Oct 19 11:28:39 2009 From: geege4 at gmail.com (geege schuman) Date: Mon, 19 Oct 2009 14:28:39 -0400 Subject: [FoRK] Ad Age In-Reply-To: <4ADC8571.2020500@lig.net> References: <493a95a00910190805ob325c9dwa2bbba663bdee47a@mail.gmail.com> <4ADC8571.2020500@lig.net> Message-ID: <493a95a00910191128t2de246c9sc16b94a8440f0684@mail.gmail.com> Very interesting and almost scary. But potentially cool too. > > Stephen > > Oddly, that's exactly what my last date said about me..... From ken_ganshirt at yahoo.ca Mon Oct 19 17:11:32 2009 From: ken_ganshirt at yahoo.ca (Ken Ganshirt @ Yahoo) Date: Mon, 19 Oct 2009 17:11:32 -0700 (PDT) Subject: [FoRK] J.Edgar's FBI returning?? Message-ID: <276112.94589.qm@web33004.mail.mud.yahoo.com> Is this worrisome? (I'm Canadian so I'm not sure .. is this as bad as the author makes it sound?) Two pages but not long. http://www.htrnews.com/article/20091015/MAN06/910150599 ...ken... __________________________________________________________________ Looking for the perfect gift? Give the gift of Flickr! http://www.flickr.com/gift/ From jbone at place.org Mon Oct 19 19:13:22 2009 From: jbone at place.org (Jeff Bone) Date: Mon, 19 Oct 2009 21:13:22 -0500 Subject: [FoRK] Skim and Plunge Message-ID: Excellent! -- http://www.stevenberlinjohnson.com/2009/09/skim-and-plunge.html SKIM AND PLUNGE The editors at Yale University Press were nice enough to invite me to edit this year's edition of Best Technology Writing. It's a great collection of essays, by some of my very favorite writers, and I encourage you to pick up a copy. I wrote an opening essay for the book that tries to wrestle with the ways in which technology writing has changed over the past few decades. Here's a section of it: The ubiquity of the digital lifestyle has forced us to write and think about technology in a different way. Think back, for example, to Stewart Brand?s classic 1973 Rolling Stone essay on the first video gamers, ?SPACEWAR: Fanatic Life and Symbolic Death Among The Computer Bums.? When Brand stumbled across those Stanford proto-gamers, battling each other via command line, it was clear to him that he?d just glimpsed the future. Of course, it took a true visionary like Brand to recognize what he?d encountered, and to write about it with such clarity and infectious curiosity--in the process inventing a whole genre of technology writing that could do justice to the encounter. But there is something about that experience that is also by definition short-sighted: any given technology will mean very different things, and have very different effects, when it is restricted to a small slice of the population. Brand?s opening line was ?Computers are coming to the people.? That was prescient enough. But as it turned out, what he saw on those screens actually had very little to do with gaming culture today. SPACEWAR let Brand sense before just about anyone else that information technology would become as mainstream as rock-and-roll or television. But he couldn?t have imagined a culture where games like Spore or Grand Theft Auto -- both of which are deftly dissected in this volume -- are far more complex, open-ended and popular than many Hollywood blockbusters. Likewise, hypertext, until mid-1994, was an emerging technology whose power users were almost all writers of experimental fiction. You could look at those links on the screen, and begin to imagine what might happen if billions of people started clicking on them. But mostly you were guessing. A shocking amount of the early commentary on hypertext-- some of it, in all honesty, written by me--focused on the radical effect hypertext would have on storytelling. Once hypertext went mainstream, however, that turned out to one of the least interesting things about it. (We?re still reading novels the old-fashioned way, one page after another.) And that?s precisely the trouble with writing about a technology when it?s still in leading indicator mode. You could look at those hyperlinks on the screen, and if you really concentrated, you might imagine a future where, say, newspaper articles linked to each other. But you could never imagine Wikipedia or YouPorn. Now we don?t have to imagine it at all: the digital future, to paraphrase William Gibson, is so much more evenly distributed among us. We don?t have to gaze into a crystal ball; we can just watch ourselves, self-reflecting as we interact with this vast new ecosystem. Some of my favorite passages in this collection have this introspective quality: the mind examining its own strange adaptation to a world that has been transformed by information technology. Consider this paragraph, from the opening section of Nicholas Carr?s ?Is Google Making Us Stupid??: I?m not thinking the way I used to think. I can feel it most strongly when I?m reading. Immersing myself in a book or a lengthy article used to be easy. My mind would get caught up in the narrative or the turns of the argument, and I?d spend hours strolling through long stretches of prose. That?s rarely the case anymore. Now my concentration often starts to drift after two or three pages. I get fidgety, lose the thread, begin looking for something else to do. I feel as if I?m always dragging my wayward brain back to the text. The deep reading that used to come naturally has become a struggle. Carr intends this as a critique, of course, and his observations will no doubt ring true for anyone who spends hours each day in front of a networked computer screen. I feel it myself right now, as I write this essay, with my open Gmail inbox hovering in the background behind the word processor, and a text message buzzing on my phone, and a whole universe of links tempting me. It is harder to sit down and focus on a linear argument or narrative for an hour at a time. In a way, our prophecies about the impact of hypertext on storytelling had it half right; it?s not that people now tell stories using branching hypertext links: it?s that we activelymiss those links when we pick up an old- fashioned book. Carr is right, too, that there is something regrettable about this shift. The kind of deep, immersive understanding that one gets from spending three hundred pages occupying another person?s consciousness is undeniably powerful and essential. And no medium rivals the book for that particular kind of thinking. But it should also be said that this kind of thinking has not simply gone away; people still read books and magazines in vast numbers. It may be harder to enter the kind of slow, contemplative state that Carr cherishes, but that doesn?t mean it?s impossible. I think of our present situation as somewhat analogous to the mass migration from the country to the city that started several centuries ago in Europe: the bustle and stimulation and diversity of urban life made it harder to enjoy the slower, organic pleasures of rural living. Still, those pleasures didn?t disappear. People continue to cherish them in mass numbers to this day. And like urban life, the new consciousness of digital culture has many benefits; it may dull certain cognitive skills, but it undoubtedly sharpens others. In his essay, Carr derides the ?skimming? habits of online readers. It?s an easy target, particularly when pitted against the hallowed activity of reading a four-hundred page novel. But skimming is an immensely valuable skill. Most of the information we interact with in our lives -- online or off -- lacks the profundity and complexity of a Great Book. We don?t need deep contemplation to assess an interoffice memo or quarterly financial report from a company we?re vaguely interested in. If we can process that information quickly and move on to more important things, so much the better. Even loftier pursuits benefit from well-developed skimming muscles. I think many of us who feel, unlike Carr, that Google has actually made us smarter operate in what I call ?skim-and-plunge? mode. We skim through pages of search results or hyperlinked articles, getting a sense of the waters, and then, when we find something interesting, we dive in and read in a slower, more engaged mode. Yes, it is probably a bit harder to become immersed in deep contemplation today than it was sitting in library in 1985, But that kind of rapid-fire skimming and discovery would have been, for all intents and purposes, impossible before the web came along. The benefits of this new consciousness go far beyond skimming of course, especially when you consider that many of the distractions are not tantalizing hyperlinks but other human beings. Here?s Andrew Sullivan describing one of the defining aspects of the experience of blogging, in his revealing essay, ?Why I Blog?: Within minutes of my posting something, even in the earliest days, readers responded. E-mail seemed to unleash their inner beast. They were more brutal than any editor, more persnickety than any copy editor, and more emotionally unstable than any colleague. Again, it?s hard to overrate how different this is... [B]efore the blogosphere, reporters and columnists were largely shielded from this kind of direct hazing. Yes, letters to the editor would arrive in due course and subscriptions would be canceled. But reporters and columnists tended to operate in a relative sanctuary, answerable mainly to their editors, not readers. For a long time, columns were essentially monologues published to applause, muffled murmurs, silence, or a distant heckle. I?d gotten blowback from pieces before?but in an amorphous, time-delayed, distant way. Now the feedback was instant, personal, and brutal. No doubt the intensity and immediacy of the feedback has its own disruptive force, making it harder for the blogger to enter the contemplative state that his forebears in the print magazine era might have enjoyed more easily. Sullivan?s description could in fact easily be marshaled in defense of Carr?s dumbing-down argument--except that where Carr sees chaos and distraction, Sullivan sees a new kind of engagement between the author and the audience. Sullivan would be the first to admit that this new kind of engagement is noisier, more offensive, and often more idiotic than any traditional interaction between author and editor. But there is so much useful signal in that noise that most of us who have sampled it find it hard to imagine going back. After all, the countryside was more polite, too. But in the end, most of us chose the city, despite all the chaos and distractions. I think we've made a similar choice with the Web today. (Excerpted from The Best Technology Writing 2009. Now go buy a copy!) From jbone at place.org Mon Oct 19 19:14:10 2009 From: jbone at place.org (Jeff Bone) Date: Mon, 19 Oct 2009 21:14:10 -0500 Subject: [FoRK] Q re: ConceptNet Message-ID: <074EEDC1-89F0-4850-BABA-49A5F3B45DDF@place.org> Anybody on list using ConceptNet or friends for anything interesting? Reasoning by analogy has some legs after all... ;-) jb From andrew at ceruleansystems.com Mon Oct 19 19:47:14 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Mon, 19 Oct 2009 19:47:14 -0700 Subject: [FoRK] J.Edgar's FBI returning?? In-Reply-To: <276112.94589.qm@web33004.mail.mud.yahoo.com> References: <276112.94589.qm@web33004.mail.mud.yahoo.com> Message-ID: <2BE6D096-E6D1-47FC-B635-F88B5E27903D@ceruleansystems.com> On Oct 19, 2009, at 5:11 PM, Ken Ganshirt @ Yahoo wrote: > Is this worrisome? (I'm Canadian so I'm not sure .. is this as bad > as the author makes it sound?) Two pages but not long. This kind of thing has been commonplace and very active since at least the mid-1990s, and probably far longer in some form. The technology leveraged is a lot more powerful than it used to be and there is a pretty strong uptick as of late, but the basic pattern is old. The pet applications vary with the administration. Whether or not it is "worrisome" is subjective and contextual. It is not a new development, though some applications are. I have no useful opinion, but this should help put it into perspective. From whump at mac.com Mon Oct 19 19:55:13 2009 From: whump at mac.com (Bill Humphries) Date: Mon, 19 Oct 2009 19:55:13 -0700 Subject: [FoRK] Ad Age In-Reply-To: <5234e09e0910190849u53ac6b94xaa2fa976b6bb9333@mail.gmail.com> References: <493a95a00910190805ob325c9dwa2bbba663bdee47a@mail.gmail.com> <5234e09e0910190849u53ac6b94xaa2fa976b6bb9333@mail.gmail.com> Message-ID: <904CAB04-64B0-4039-9136-6EA254BEF31E@mac.com> On Oct 19, 2009, at 8:49 AM, HK Pang wrote: > So, can facebook legally pass this piece of info to advertisers? In the US, yes. And it's annoying. When a friend of mine got tired of facebook demanding to know her gender presentation, she was barraged with "You Are Fat!" ad nonsense. Before that, she got ads for developer tools and Dice. -- whump From andrew at ceruleansystems.com Mon Oct 19 20:06:20 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Mon, 19 Oct 2009 20:06:20 -0700 Subject: [FoRK] Q re: ConceptNet In-Reply-To: <074EEDC1-89F0-4850-BABA-49A5F3B45DDF@place.org> References: <074EEDC1-89F0-4850-BABA-49A5F3B45DDF@place.org> Message-ID: <0E424152-0D51-4CD1-9D38-731108EBB0DF@ceruleansystems.com> On Oct 19, 2009, at 7:14 PM, Jeff Bone wrote: > > Anybody on list using ConceptNet or friends for anything interesting? > > Reasoning by analogy has some legs after all... ;-) Are we talking about ConceptNet/OpenMind specifically or semantic web technologies generally? And how do you define "interesting"? The stuff you can download off the web doesn't scale much beyond toys but a lot R&D dollars are being spent on the topic. From whump at mac.com Mon Oct 19 20:09:06 2009 From: whump at mac.com (Bill Humphries) Date: Mon, 19 Oct 2009 20:09:06 -0700 Subject: [FoRK] Ad Age In-Reply-To: <904CAB04-64B0-4039-9136-6EA254BEF31E@mac.com> References: <493a95a00910190805ob325c9dwa2bbba663bdee47a@mail.gmail.com> <5234e09e0910190849u53ac6b94xaa2fa976b6bb9333@mail.gmail.com> <904CAB04-64B0-4039-9136-6EA254BEF31E@mac.com> Message-ID: On Oct 19, 2009, at 7:55 PM, Bill Humphries wrote: > > On Oct 19, 2009, at 8:49 AM, HK Pang wrote: > >> So, can facebook legally pass this piece of info to advertisers? > > In the US, yes. Let me clarify. Facebook can go to its inventory of ads and say, show an ad targeted for a 45 yo married, heterosexual, white woman living in 94011 + other refinements. The advertiser buys impressions and tells Facebook who they want to target. The advertiser may not be as interested in the identity of who's getting the impression, as they are in getting the impression in front of what they consider the niche. I don't know what info they pass back up on impressions and clickthroughs. But facebook's selling point is highly targeted profile ad exposure, so the less info you give them, the less valuable you are to them as a pair of eyeballs, hence the dunning to give up more info like gender presentation. However, as my example earlier points out, the targeting can be biased by stereotypes about gender. -- whump From dmorton at bitfurnace.com Mon Oct 19 21:30:46 2009 From: dmorton at bitfurnace.com (Damien Morton) Date: Tue, 20 Oct 2009 15:30:46 +1100 Subject: [FoRK] Ad Age In-Reply-To: References: <493a95a00910190805ob325c9dwa2bbba663bdee47a@mail.gmail.com> <5234e09e0910190849u53ac6b94xaa2fa976b6bb9333@mail.gmail.com> <904CAB04-64B0-4039-9136-6EA254BEF31E@mac.com> Message-ID: <8092dc770910192130rd7e1388i7118966af469cb9@mail.gmail.com> On Tue, Oct 20, 2009 at 2:09 PM, Bill Humphries wrote: > > > Let me clarify. Facebook can go to its inventory of ads and say, show an ad > targeted for a 45 yo married, heterosexual, white woman living in 94011 + > other refinements. The advertiser buys impressions and tells Facebook who > they want to target. The advertiser may not be as interested in the identity > of who's getting the impression, as they are in getting the impression in > front of what they consider the niche. > > I don't know what info they pass back up on impressions and clickthroughs. > > Theoretically, what you describe is possible, however, when I attempted something like this - looking for iPhone owners who are into fishing and related sports, their system completely failed me. Essentially all they provide is a keyword search on their mound of data. Your "45 yo married, heterosexual, white woman living in 94011 + other refinements." targeting isnt currently possible using their system. From ken_ganshirt at yahoo.ca Mon Oct 19 22:07:49 2009 From: ken_ganshirt at yahoo.ca (Ken Ganshirt @ Yahoo) Date: Mon, 19 Oct 2009 22:07:49 -0700 (PDT) Subject: [FoRK] Ad Age In-Reply-To: <8092dc770910192130rd7e1388i7118966af469cb9@mail.gmail.com> Message-ID: <181503.28588.qm@web33002.mail.mud.yahoo.com> --- On Mon, 10/19/09, Damien Morton wrote: > Theoretically, what you describe is possible, however, when I attempted > something like this - looking for iPhone owners who are into fishing and > related sports, their system completely failed me. > > Essentially all they provide is a keyword search on their mound of data. > Your "45 yo married, heterosexual, white woman living in 94011 + other > refinements." targeting isnt currently possible using their > system. > Not to you, of course. Certainly not for free. But I'm sure they have access to that information quite easily. ...ken... __________________________________________________________________ Yahoo! Canada Toolbar: Search from anywhere on the web, and bookmark your favourite sites. Download it now http://ca.toolbar.yahoo.com. From sdw at lig.net Mon Oct 19 23:04:48 2009 From: sdw at lig.net (Stephen Williams) Date: Mon, 19 Oct 2009 23:04:48 -0700 Subject: [FoRK] J.Edgar's FBI returning?? In-Reply-To: <2BE6D096-E6D1-47FC-B635-F88B5E27903D@ceruleansystems.com> References: <276112.94589.qm@web33004.mail.mud.yahoo.com> <2BE6D096-E6D1-47FC-B635-F88B5E27903D@ceruleansystems.com> Message-ID: <4ADD5300.2030305@lig.net> J. Andrew Rogers wrote: > > On Oct 19, 2009, at 5:11 PM, Ken Ganshirt @ Yahoo wrote: >> Is this worrisome? (I'm Canadian so I'm not sure .. is this as bad as >> the author makes it sound?) Two pages but not long. > > > This kind of thing has been commonplace and very active since at least > the mid-1990s, and probably far longer in some form. The technology > leveraged is a lot more powerful than it used to be and there is a > pretty strong uptick as of late, but the basic pattern is old. The pet > applications vary with the administration. > > Whether or not it is "worrisome" is subjective and contextual. It is > not a new development, though some applications are. I have no useful > opinion, but this should help put it into perspective. It is not unusual, however it should be strictly controlled and the information obtained only usable for certain purposes, such as anti-terrorism. And the watchers need to be watched. Really, if "they" get this right, we the people should have a corresponding right to watch them. Unfortunately, many interesting ways of watching them would be considered highly illegal. Some people have no sense of humor. Congress should be overseeing them with vigor to make sure that nothing untoward slips by. That was not the case under the GOP, at least. I really liked the old FCC rule, superseded by the damn cell companies, that you could receive and decode any signal you were capable of, but that there were restrictions on what you did with official traffic. It is a little weird for the government to know, say, all about your sex life or whatever. On the other hand, if the privacy controls are more or less equivalent to a compartmentalized "top secret" classification, unless and until you are a confirmed terrorist threat, it's not so bad. As long as you can extract severe penalties for abuse and loss and there is protection against cover up, it is a relatively safe system. It should be no worse than your doctor or attorney knowing something. Not that we are there yet. In reality, most people in government are very good people. A bad actor is probably going to be found out by his coworkers long before a typical citizen suffers from corruption. Notwithstanding a high error rate in the "last mile" of law enforcement or overseas assignments with contractor security companies. Stephen From whump at mac.com Mon Oct 19 23:17:03 2009 From: whump at mac.com (Bill Humphries) Date: Mon, 19 Oct 2009 23:17:03 -0700 Subject: [FoRK] Ad Age In-Reply-To: <8092dc770910192130rd7e1388i7118966af469cb9@mail.gmail.com> References: <493a95a00910190805ob325c9dwa2bbba663bdee47a@mail.gmail.com> <5234e09e0910190849u53ac6b94xaa2fa976b6bb9333@mail.gmail.com> <904CAB04-64B0-4039-9136-6EA254BEF31E@mac.com> <8092dc770910192130rd7e1388i7118966af469cb9@mail.gmail.com> Message-ID: <5AFD7F6C-D344-4089-85D0-49EE9DA7F754@mac.com> On Oct 19, 2009, at 9:30 PM, Damien Morton wrote: > Essentially all they provide is a keyword search on their mound of > data. > Your "45 yo married, heterosexual, white woman living in 94011 + other > refinements." targeting isnt currently possible using their system. Ah, it's not as refined as I thought. I guess I can take comfort in that. Finding out what marketers want to sell to my thinly sliced demographic might be more than I want to know. -- whump From wgstoddard at gmail.com Wed Oct 21 05:25:23 2009 From: wgstoddard at gmail.com (Bill Stoddard) Date: Wed, 21 Oct 2009 08:25:23 -0400 Subject: [FoRK] For the arm chair macro economists on FoRK Message-ID: <4ADEFDB3.6070109@gmail.com> A very good blog worth following: http://mpettis.com/ From the website: Michael Pettis is a professor at Peking University?s Guanghua School of Management, where he specializes in Chinese financial markets, and a Senior Associate at the Carnegie Endowment for International Peace. Latest post has many good bits with geopolitical/macroeconomic implications: http://mpettis.com/2009/10/china%E2%80%99s-september-data-suggest-that-the-long-term-overcapacity-problem-is-only-intensifying/ Bill From jbone at place.org Wed Oct 21 07:43:43 2009 From: jbone at place.org (Jeff Bone) Date: Wed, 21 Oct 2009 09:43:43 -0500 Subject: [FoRK] Q re: ConceptNet (also FluidDB) Message-ID: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> JAR asks: > Are we talking about ConceptNet/OpenMind specifically or semantic > web technologies generally? And how do you define "interesting"? ConceptNet specifically. It's got some interesting attributes and applications that most of the "strong" ontological systems don't have (at least in my own uses of them.) "Interesting" in this case means just that --- useful beyond what e.g. Cyc, various rdf / semweb technologies, and so on have been in my experience. Particularly when dealing with large and diverse natural language corpuses, CN2/3 (with supplemental data) have proven more useful at various extraction, classification, and extrapolation tasks than other methods including statistical ones, naive bayes, etc. (For me. YMMV.) The actual CN tools themselves aren't as useful (you're correct, toys so far) as the knowledge base per se and its data model --- which can be easily embedded into a slightly more robust model that more easily and effectively handles meta-information such as provenance, etc. --- and does so defeasibly if necessary, important for real-world use. IMHO, those are the major conceptual problems w/ the rdf-like approaches; representing defeasible information as-such and handling reification and self-referentiality. Do-able, but not in a satisfying or particularly practical way. Other problems include the mapping problem, while CN's natural language fragment / semi-structured approach works much better for fast-and-loose reasoning over mostly- raw texts -- e.g. stretching to various machine learning goals for which strong induction and inference engines aren't really well-suited. Other interesting bits along tangential lines: FreeBase and its more- structured Wikipedia extract (WEX) have been proving useful, but *this* --- aside from some of the intended monolithic web-app properties --- is *really* interesting from a data model perspective: http://www.fluidinfo.com/ In particular check out the talks and articles: http://www.fluidinfo.com/talks jb From eugen at leitl.org Wed Oct 21 09:22:15 2009 From: eugen at leitl.org (Eugen Leitl) Date: Wed, 21 Oct 2009 18:22:15 +0200 Subject: [FoRK] Dominance, Politics, and Physiology: Voters' Testosterone Changes on the Night of the 2008 United States Presidential Election Message-ID: <20091021162215.GC27331@leitl.org> http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0007543 Dominance, Politics, and Physiology: Voters' Testosterone Changes on the Night of the 2008 United States Presidential Election Steven J. Stanton1, Jacinta C. Beehner2,3, Ekjyot K. Saini3, Cynthia M. Kuhn4, Kevin S. LaBar1* 1 Center for Cognitive Neuroscience, Duke University, Durham, North Carolina, United States of America, 2 Department of Anthropology, University of Michigan, Ann Arbor, Michigan, United States of America, 3 Department of Psychology, University of Michigan, Ann Arbor, Michigan, United States of America, 4 Department of Pharmacology and Cancer Biology, Duke University, Durham, North Carolina, United States of America Abstract Top Background Political elections are dominance competitions. When men win a dominance competition, their testosterone levels rise or remain stable to resist a circadian decline; and when they lose, their testosterone levels fall. However, it is unknown whether this pattern of testosterone change extends beyond interpersonal competitions to the vicarious experience of winning or losing in the context of political elections. Women's testosterone responses to dominance competition outcomes are understudied, and to date, a clear pattern of testosterone changes in response to winning and losing dominance competitions has not emerged. Methodology/Principal Findings The present study investigated voters' testosterone responses to the outcome of the 2008 United States Presidential election. 183 participants provided multiple saliva samples before and after the winner was announced on Election Night. The results show that male Barack Obama voters (winners) had stable post-outcome testosterone levels, whereas testosterone levels dropped in male John McCain and Robert Barr voters (losers). There were no significant effects in female voters. Conclusions/Significance The findings indicate that male voters exhibit biological responses to the realignment of a country's dominance hierarchy as if they participated in an interpersonal dominance contest. Citation: Stanton SJ, Beehner JC, Saini EK, Kuhn CM, LaBar KS (2009) Dominance, Politics, and Physiology: Voters' Testosterone Changes on the Night of the 2008 United States Presidential Election. PLoS ONE 4(10): e7543. doi:10.1371/journal.pone.0007543 Editor: Colin Allen, Indiana University, United States of America Received: July 20, 2009; Accepted: September 28, 2009; Published: October 21, 2009 Copyright: ? 2009 Stanton et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This research was supported by departmental funds from Duke University (to KSL) and the University of Michigan (to JCB) and a McClelland Postdoctoral Fellowship (to SJS). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. * E-mail: klabar at duke.edu -- Eugen* Leitl leitl http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE From andrew at ceruleansystems.com Wed Oct 21 10:03:40 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Wed, 21 Oct 2009 10:03:40 -0700 Subject: [FoRK] Q re: ConceptNet (also FluidDB) In-Reply-To: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> References: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> Message-ID: On Oct 21, 2009, at 7:43 AM, Jeff Bone wrote: > JAR asks: > >> Are we talking about ConceptNet/OpenMind specifically or semantic >> web technologies generally? And how do you define "interesting"? > > ConceptNet specifically. It's got some interesting attributes and > applications that most of the "strong" ontological systems don't > have (at least in my own uses of them.) "Interesting" in this case > means just that --- useful beyond what e.g. Cyc, various rdf / > semweb technologies, and so on have been in my experience. > Particularly when dealing with large and diverse natural language > corpuses, CN2/3 (with supplemental data) have proven more useful at > various extraction, classification, and extrapolation tasks than > other methods including statistical ones, naive bayes, etc. (For > me. YMMV.) In my experience, you are entirely correct that most semantic web technologies are over-structured to the point of being not very useful. Most of the interesting R&D currently is on very general graph analytic systems that subsume the rigid classical models but do not require them. Rigid models are more susceptible to the numerous NP land mines that litter this theoretical landscape. > The actual CN tools themselves aren't as useful (you're correct, > toys so far) as the knowledge base per se and its data model --- > which can be easily embedded into a slightly more robust model that > more easily and effectively handles meta-information such as > provenance, etc. --- and does so defeasibly if necessary, important > for real-world use. IMHO, those are the major conceptual problems > w/ the rdf-like approaches; representing defeasible information as- > such and handling reification and self-referentiality. Do-able, but > not in a satisfying or particularly practical way. Yes, which is why a lot of the current R&D is focused on generalized graph-like computation, not the narrow case of RDF-like systems. You can do it with RDF-like systems in principle, but not efficiently and efficiency is very important for most real work at the scales required. Current popular tools and models are badly designed for the actual markets for this kind of technology. The primary real (and "interesting") use case for these types of models in commercial and other systems is induction and prediction in highly dynamic data environments at large scales. There are organizations starting to build more generalized graph systems for this purpose at very large scales, but it definitely isn't open source (or even shrink-wrap) technology at this point. It will probably be a few years before this creeps into the web at large. From jbone at place.org Wed Oct 21 11:00:32 2009 From: jbone at place.org (Jeff Bone) Date: Wed, 21 Oct 2009 13:00:32 -0500 Subject: [FoRK] Q re: ConceptNet (also FluidDB) In-Reply-To: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> References: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> Message-ID: Re: JAR's comments... Yup to all, particularly the graph theoretical stuff. Important to note (of course you realize this already, just noting "for the record) that rdf / semantic networks, your usual induction and inference engines, and so forth are all just specialized graph engines with some strict constraints on the meanings of nodes and edges and the algorithms which operate on them. IMHO the innovation in ConceptNet is in having a higher-order / fuzzier semantics for the nodes and edges. CN's concepts and relations operate at a level a lot closer to the fuzzy semantic / analogical reasoning of humans (and hence the real uses of natural language) than e.g. the stricter systems. But at some level it's all graph theory, whether we're talking CN, rdf / cwm / SparQL / etc., Prolog-like systems, and so on. Neural networks and other similar systems are something else entirely, though, and while there's a mapping here it's a bit elusive. Spreading activation in semantic networks with fuzzy, defeasible semantics seems like a pretty rich topic at present. jb From sdw at lig.net Wed Oct 21 12:24:25 2009 From: sdw at lig.net (Stephen Williams) Date: Wed, 21 Oct 2009 12:24:25 -0700 Subject: [FoRK] Q re: ConceptNet (also FluidDB) In-Reply-To: References: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> Message-ID: <4ADF5FE9.6030501@lig.net> Jeff Bone wrote: > > Re: JAR's comments... > > Yup to all, particularly the graph theoretical stuff. Important to > note (of course you realize this already, just noting "for the record) > that rdf / semantic networks, your usual induction and inference > engines, and so forth are all just specialized graph engines with some > strict constraints on the meanings of nodes and edges and the > algorithms which operate on them. IMHO the innovation in ConceptNet > is in having a higher-order / fuzzier semantics for the nodes and > edges. CN's concepts and relations operate at a level a lot closer to > the fuzzy semantic / analogical reasoning of humans (and hence the > real uses of natural language) than e.g. the stricter systems. I view a lot of the semantic technology as A) a better way to organize information for straight business purposes and B) often the best form of bottom-up feeder information that could feed into knowledge and probabilistic graph reasoning systems. When you can't or don't need to capture precise probabilistic knowledge, RDF et al is just fine and far simpler. Currently however, most tools are annoyingly formal and stiff. We need projects like ConceptNet to make it easy to capture and use knowledge, hopefully spurring multiple hot applications that get a good feedback loop going. > > But at some level it's all graph theory, whether we're talking CN, rdf > / cwm / SparQL / etc., Prolog-like systems, and so on. Exactly. > > Neural networks and other similar systems are something else entirely, > though, and while there's a mapping here it's a bit elusive. > Spreading activation in semantic networks with fuzzy, defeasible > semantics seems like a pretty rich topic at present. I'm glad you now see a mapping / equivalence. I would really like to focus on proving what my intuition tells me, but have several more pressing concerns first. Someone will eventually map the equivalence of NN with probabilistic relational graph reasoning (Bayesian / Markov with imprecise reasoning, etc.) and both will be A) unified or B) both be stronger as a result. sdw > > jb From sdw at lig.net Wed Oct 21 17:52:23 2009 From: sdw at lig.net (Stephen Williams) Date: Wed, 21 Oct 2009 17:52:23 -0700 Subject: [FoRK] Redox battery for EVs is recharged in minutes through electrolyte exchange Message-ID: <4ADFACC7.7010506@lig.net> Cool. As long as we don't get exploding / lightning-capable liquid at the GAS STATION. http://www.edn.com/blog/1470000147/post/70049807.html Redox battery for EVs is recharged in minutes through electrolyte exchange Oct 15 2009 4:37PM | Permalink |Comments (7) | In general, batteries are good at storing energy and not so good at charging or discharging rapidly. Supercaps, on the other hand, are excellent at fast charging/discharging, but not so good at storing energy over a period of days or months. Ideally, for energy storage and delivery for the next generation of electric vehicles, you need both characteristics. After all, it takes just 5 minutes to gas (or diesel) up a conventional car. The 4 ? 10 hours it takes to re-charge today?s batteries is a sticking point for EVs of the future. Frounhofer prototype EV w/redox batteryThe redox battery (which is shorthand for reduction-oxidation flow battery) offers a novel solution to the problem of charging times: Replace the discharged electrolyte with a fully-charged electrolyte much as you?d fill up an empty tank with more gasoline. Here?s the Wikipeida description: ?A flow battery is a form of rechargeable battery in which electrolyte containing one or more dissolved electroactive species flows through an electrochemical cell that converts chemical energy directly to electricity? Flow batteries can be rapidly "recharged" by replacing the electrolyte liquid (in a similar way to refilling fuel tanks for internal combustion engines) while simultaneously recovering the spent material for re-energization.? The problem with redox batteries is that they can?t store as much energy as, say, a lithium ion battery ?only a quarter as much. However, researchers at the Fraunhofer Institute for Chemical Technology ICT have refined the process for a redox cell that allows a four-to five-fold increase in energy storage, making it on par with lithium ion batteries. (No details so far on the exact refinements.) The German government is aiming for one million electric cars being sold in Germany by the year 2020. From andrew at ceruleansystems.com Wed Oct 21 20:38:39 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Wed, 21 Oct 2009 20:38:39 -0700 Subject: [FoRK] Q re: ConceptNet (also FluidDB) In-Reply-To: <4ADF5FE9.6030501@lig.net> References: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> <4ADF5FE9.6030501@lig.net> Message-ID: <8F8BAAA4-298A-4E49-9C84-42C45EAA52CD@ceruleansystems.com> On Oct 21, 2009, at 12:24 PM, Stephen Williams wrote: > I view a lot of the semantic technology as A) a better way to > organize information for straight business purposes and B) often the > best form of bottom-up feeder information that could feed into > knowledge and probabilistic graph reasoning systems. A (the?) big problem is that even if A) is true in theory it is not true in practice. The organization exposed to the user has little relation to the organization under the hood. That impedance mismatch is usually very expensive. Explicitly structured data typically matches common use cases pretty well, a cheap "80/20 rule" optimization. > I'm glad you now see a mapping / equivalence. I would really like > to focus on proving what my intuition tells me, but have several > more pressing concerns first. Someone will eventually map the > equivalence of NN with probabilistic relational graph reasoning > (Bayesian / Markov with imprecise reasoning, etc.) and both will be > A) unified or B) both be stronger as a result. A lot of work gets done that makes the assumption that classic NN models are inferior approximations of high-order Bayesian models. It may just be a conjecture that is commonly assumed, though it does appear to be obviously true upon non-rigorous inspection. Anyone can add weights to the edges and vertices of a graph, the real trick is deciding where to not put edges even when you have a nominal reason to. I've long wondered if there is a relationship between Braess' Paradox and pruning algorithms for graph-like computational models but am too busy/lazy to spend much time thinking about it. From eugen at leitl.org Thu Oct 22 01:24:14 2009 From: eugen at leitl.org (Eugen Leitl) Date: Thu, 22 Oct 2009 10:24:14 +0200 Subject: [FoRK] Redox battery for EVs is recharged in minutes through electrolyte exchange In-Reply-To: <4ADFACC7.7010506@lig.net> References: <4ADFACC7.7010506@lig.net> Message-ID: <20091022082414.GR27331@leitl.org> On Wed, Oct 21, 2009 at 05:52:23PM -0700, Stephen Williams wrote: > Cool. As long as we don't get exploding / lightning-capable liquid at > the GAS STATION. I have this neat idea. Why not taking liquid methanol, which has half the energy density of diesel/gasoline, and use this as one electrode in a redox pair, the other being air. You don't have to carry the oxidant, and you lose mass as one electrode is consumed, since you vent the reaction products (water and CO2) into air. I think I will call it a "fuel cell". And since it's running with straight methanol, I'll call it a "direct methanol fuel cell" or DMFC. Do you think I can patent this? > http://www.edn.com/blog/1470000147/post/70049807.html > > Redox battery for EVs is recharged in minutes through electrolyte exchange -- Eugen* Leitl leitl http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE From sdw at lig.net Thu Oct 22 02:07:27 2009 From: sdw at lig.net (Stephen Williams) Date: Thu, 22 Oct 2009 02:07:27 -0700 Subject: [FoRK] Redox battery for EVs is recharged in minutes through electrolyte exchange In-Reply-To: <20091022082414.GR27331@leitl.org> References: <4ADFACC7.7010506@lig.net> <20091022082414.GR27331@leitl.org> Message-ID: <4AE020CF.9090603@lig.net> Eugen Leitl wrote: > On Wed, Oct 21, 2009 at 05:52:23PM -0700, Stephen Williams wrote: > >> Cool. As long as we don't get exploding / lightning-capable liquid at >> the GAS STATION. >> > > I have this neat idea. Why not taking liquid methanol, which > has half the energy density of diesel/gasoline, and use this > as one electrode in a redox pair, the other being air. You > don't have to carry the oxidant, and you lose mass as one > electrode is consumed, since you vent the reaction products > (water and CO2) into air. I think I will call it a "fuel cell". > And since it's running with straight methanol, I'll call > it a "direct methanol fuel cell" or DMFC. Do you think I can > patent this? > You didn't list the other key component on your patent application: What is this boxed marked "MOH [1]" where electricity goes in at the "gas" station and methanol comes out? One of the main problems with an all-electric vehicle is fast charging, or charging at all. Swapping out a liquid battery is clever, if you can get it to work. Any fuel cell fuel is great if you can make it and get it to "gas" stations more efficiently than petrol. [This spot for a wild and crazy idea.]: Create tracking intense light projecting zones on streets, highways, and especially hills. Use grates, mirrors, whatever. As photo-rechargeable cars drive over the charging zones, focused, high intensity light illuminates just the receptors on the bottom of cars. The intense pulse is captured by multi-layer / fractal / nano receivers and the impulse is stored in capacitor banks that slowly drain into the batteries / wheels. Presto: an optical (or microwave, or even inductive) "third rail". Probable power scale mismatch, probably safety issues... [1] MOH = Magic occurs here. sdw > > >> http://www.edn.com/blog/1470000147/post/70049807.html >> >> Redox battery for EVs is recharged in minutes through electrolyte exchange >> > > From sdw at lig.net Thu Oct 22 02:21:54 2009 From: sdw at lig.net (Stephen Williams) Date: Thu, 22 Oct 2009 02:21:54 -0700 Subject: [FoRK] Q re: ConceptNet (also FluidDB) In-Reply-To: <8F8BAAA4-298A-4E49-9C84-42C45EAA52CD@ceruleansystems.com> References: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> <4ADF5FE9.6030501@lig.net> <8F8BAAA4-298A-4E49-9C84-42C45EAA52CD@ceruleansystems.com> Message-ID: <4AE02432.3010906@lig.net> J. Andrew Rogers wrote: > > On Oct 21, 2009, at 12:24 PM, Stephen Williams wrote: >> I view a lot of the semantic technology as A) a better way to >> organize information for straight business purposes and B) often the >> best form of bottom-up feeder information that could feed into >> knowledge and probabilistic graph reasoning systems. > > > A (the?) big problem is that even if A) is true in theory it is not > true in practice. The organization exposed to the user has little > relation to the organization under the hood. That impedance mismatch > is usually very expensive. Explicitly structured data typically > matches common use cases pretty well, a cheap "80/20 rule" optimization. > There was a commercial database product with a very nice Java-based GUI that was essentially an RDF-graph based database. It was not hard to understand or use. Unfortunately, they took the wrong product path. I had several conversations with the founder. He just deviate from his particular viewpoint and he didn't know anything about RDF / semantic ideas. Anyway, yes, that is true now, however I don't think it is necessarily true. We just need the equivalent of a spreadsheet application to make it accessible. I'm thinking about it, when I can... > >> I'm glad you now see a mapping / equivalence. I would really like to >> focus on proving what my intuition tells me, but have several more >> pressing concerns first. Someone will eventually map the equivalence >> of NN with probabilistic relational graph reasoning (Bayesian / >> Markov with imprecise reasoning, etc.) and both will be A) unified or >> B) both be stronger as a result. > > > A lot of work gets done that makes the assumption that classic NN > models are inferior approximations of high-order Bayesian models. It > may just be a conjecture that is commonly assumed, though it does > appear to be obviously true upon non-rigorous inspection. Anyone can > add weights to the edges and vertices of a graph, the real trick is > deciding where to not put edges even when you have a nominal reason to. > > > I've long wondered if there is a relationship between Braess' Paradox > and pruning algorithms for graph-like computational models but am too > busy/lazy to spend much time thinking about it. Seems like a different thing to me... Although I wouldn't be surprised if there is some analog. Stephen From eugen at leitl.org Thu Oct 22 02:45:13 2009 From: eugen at leitl.org (Eugen Leitl) Date: Thu, 22 Oct 2009 11:45:13 +0200 Subject: [FoRK] Redox battery for EVs is recharged in minutes through electrolyte exchange In-Reply-To: <4AE020CF.9090603@lig.net> References: <4ADFACC7.7010506@lig.net> <20091022082414.GR27331@leitl.org> <4AE020CF.9090603@lig.net> Message-ID: <20091022094513.GT27331@leitl.org> On Thu, Oct 22, 2009 at 02:07:27AM -0700, Stephen Williams wrote: > You didn't list the other key component on your patent application: What > is this boxed marked "MOH [1]" where electricity goes in at the "gas" > station and methanol comes out? Unfortunately, that is already accounted for: http://en.wikipedia.org/wiki/Methanol Sadly, very little magic involved. Extra bonus points for making a direct methane oxidation to methanol microreactor for your cellar though, running off the natural gas pipe. But wait, if you have a mild methanol electrosynthesis process from air or flue CO2 and water, why not not putting that box at home, and run it off your roof PV surplus? What do you need a gas station for, again? > One of the main problems with an all-electric vehicle is fast charging, Why do you need fast charging, if your car is parked in the garage at home or in the company parking lot? Billions of EUR are being spent right now installing charging stations everywhere in France alone. > or charging at all. Swapping out a liquid battery is clever, if you can A liquid battery which has <0.23 MJ/kg density (diesel: 38.6 MJ/l, gasoline: 34.2 MJ/l, methanol 15.6 MJ/l, Li-ion 0.8-0.9 MJ/kg, this liquid thing, purportedly one quarter of that) and needs to be pumped out and recycled in nonexisting infrastructure is the opposite of clever. It's hard enough to retrofit existing gas stations for a methanol pump. > get it to work. Any fuel cell fuel is great if you can make it and get > it to "gas" stations more efficiently than petrol. Either you make it onsite from natural gas, or via photovoltaics, or you wheel it in the regular way. You just need a tank and a pumping station which is methanol-certified. > [This spot for a wild and crazy idea.]: > Create tracking intense light projecting zones on streets, highways, and > especially hills. Use grates, mirrors, whatever. As photo-rechargeable > cars drive over the charging zones, focused, high intensity light > illuminates just the receptors on the bottom of cars. The intense pulse > is captured by multi-layer / fractal / nano receivers and the impulse is > stored in capacitor banks that slowly drain into the batteries / > wheels. Presto: an optical (or microwave, or even inductive) "third > rail". Probable power scale mismatch, probably safety issues... Or you could just use a PV-driven power rail, with automatic trolleys. > [1] MOH = Magic occurs here. I much prefer MeOH to MOH. -- Eugen* Leitl leitl http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE From dave.long at bluewin.ch Thu Oct 22 05:02:59 2009 From: dave.long at bluewin.ch (Dave Long) Date: Thu, 22 Oct 2009 14:02:59 +0200 Subject: [FoRK] Q re: ConceptNet (also FluidDB) In-Reply-To: <4AE02432.3010906@lig.net> References: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> <4ADF5FE9.6030501@lig.net> <8F8BAAA4-298A-4E49-9C84-42C45EAA52CD@ceruleansystems.com> <4AE02432.3010906@lig.net> Message-ID: >> Braess' Paradox btw, i've found figures 1 and 4 from the following helpful in getting some intuition for this... http://www.rockefeller.edu/labheads/cohenje/PDFs/ 185CohenHorowitzNature1991.pdf -Dave From howell.r at inkworkswell.com Thu Oct 22 07:02:39 2009 From: howell.r at inkworkswell.com (Reese) Date: Thu, 22 Oct 2009 09:02:39 -0500 Subject: [FoRK] Nook E-Reader - or Nookie Reader Message-ID: <4AE065FF.1000704@inkworkswell.com> I like the one comment on this device. Reese -------- Original Message -------- Subject: Google Alert - Date: Wed, 21 Oct 2009 20:26:14 +0000 Barnes & Noble Announces ?Nook? E-Book Reader. X-bit Labs Unfortunately, Nook still has two major disadvantages of Amazon Kindle: it works and is available only in the US (there is an international version of ... From meltsner at alum.mit.edu Thu Oct 22 06:18:36 2009 From: meltsner at alum.mit.edu (Ken Meltsner) Date: Thu, 22 Oct 2009 08:18:36 -0500 Subject: [FoRK] Nook E-Reader - or Nookie Reader In-Reply-To: <4AE065FF.1000704@inkworkswell.com> References: <4AE065FF.1000704@inkworkswell.com> Message-ID: <692a81590910220618k27e21688gb16e2a1ef02837a0@mail.gmail.com> Even if/when the Kindle and its ilk work internationally, there are still lots of legal entanglements with international copyright and licensing. For example, the Kindle has certain books that are only available in specific countries because the publisher doesn't have the rights in other countries. Just one more example of the 21st century clashing with the 19th. Ken Meltsner From eugen at leitl.org Thu Oct 22 06:22:31 2009 From: eugen at leitl.org (Eugen Leitl) Date: Thu, 22 Oct 2009 15:22:31 +0200 Subject: [FoRK] Nook E-Reader - or Nookie Reader In-Reply-To: <692a81590910220618k27e21688gb16e2a1ef02837a0@mail.gmail.com> References: <4AE065FF.1000704@inkworkswell.com> <692a81590910220618k27e21688gb16e2a1ef02837a0@mail.gmail.com> Message-ID: <20091022132231.GZ27331@leitl.org> On Thu, Oct 22, 2009 at 08:18:36AM -0500, Ken Meltsner wrote: > Just one more example of the 21st century clashing with the 19th. Nothing what bookchan can't handle. -- Eugen* Leitl leitl http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE From howell.r at inkworkswell.com Thu Oct 22 08:49:46 2009 From: howell.r at inkworkswell.com (Reese) Date: Thu, 22 Oct 2009 10:49:46 -0500 Subject: [FoRK] Redox battery for EVs is recharged in minutes through electrolyte exchange In-Reply-To: <20091022082414.GR27331@leitl.org> References: <4ADFACC7.7010506@lig.net> <20091022082414.GR27331@leitl.org> Message-ID: <4AE07F1A.9010800@inkworkswell.com> Eugen Leitl wrote: > On Wed, Oct 21, 2009 at 05:52:23PM -0700, Stephen Williams wrote: >> Cool. As long as we don't get exploding / lightning-capable liquid at >> the GAS STATION. > > I have this neat idea. Why not taking liquid methanol, which > has half the energy density of diesel/gasoline, and use this > as one electrode in a redox pair, the other being air. You > don't have to carry the oxidant, and you lose mass as one > electrode is consumed, since you vent the reaction products > (water and CO2) into air. I think I will call it a "fuel cell". Given everything else that is going on with CO2, why would you want to vent more of it to the atmosphere? Reese From eugen at leitl.org Thu Oct 22 08:36:58 2009 From: eugen at leitl.org (Eugen Leitl) Date: Thu, 22 Oct 2009 17:36:58 +0200 Subject: [FoRK] Redox battery for EVs is recharged in minutes through electrolyte exchange In-Reply-To: <4AE07F1A.9010800@inkworkswell.com> References: <4ADFACC7.7010506@lig.net> <20091022082414.GR27331@leitl.org> <4AE07F1A.9010800@inkworkswell.com> Message-ID: <20091022153658.GC27331@leitl.org> On Thu, Oct 22, 2009 at 10:49:46AM -0500, Reese wrote: > >I have this neat idea. Why not taking liquid methanol, which > >has half the energy density of diesel/gasoline, and use this > >as one electrode in a redox pair, the other being air. You > >don't have to carry the oxidant, and you lose mass as one > >electrode is consumed, since you vent the reaction products > >(water and CO2) into air. I think I will call it a "fuel cell". > > Given everything else that is going on with CO2, why would you want > to vent more of it to the atmosphere? Because an EV scales into weight regions which no ICE can reach, since it can have regenerative braking, spike cache, in-wheel motors and similar. Nevermind that ICE cars have about 23% efficiency which can't be raised, while fuel cells are non-Carnot, so DMFCs can go to 40% or higher (the methanol permeability of the proton membrane has been fixed in the lab with multiple nanocoatings, not yet in practice though). So you burn way less fuel and at a higher efficiency. Plus MeOH has more hydrogen than alcanes/aromates in gasoline/diesel, can be made from methane (including direct methane oxidation) or renewables, so net CO2 can be neutral. E.g. http://www.physorg.com/news159098987.html but presumably you can drive this photonically, or via a electrosynthetic route, presumably scaling down to very small facility scale, down to fridge-sized units. Not that motor vehicles are the main culprit, in 2006 it was some 33% of total CO2 emissions in the US. You can shrink this to below 10% if you'd just substitute them with DMFCs. Don't get me wrong, EVs will be there eventually, but methanol is an excellent bridge technology to both EVs and hydrogen. See http://www.amazon.com/Beyond-Oil-Gas-Methanol-Economy/dp/3527324224/ why this is so (Olah et al. are not entirely correct there though, caveat lector). -- Eugen* Leitl leitl http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE From saigua at sbcglobal.net Thu Oct 22 11:40:00 2009 From: saigua at sbcglobal.net (Steve Nordquist) Date: Thu, 22 Oct 2009 13:40:00 -0500 Subject: [FoRK] Ad Age In-Reply-To: <493a95a00910190805ob325c9dwa2bbba663bdee47a@mail.gmail.com> References: <493a95a00910190805ob325c9dwa2bbba663bdee47a@mail.gmail.com> Message-ID: > http://adage.com/mediaworks/article?article_id=139760 > > third-fastest-growing economic sector over the next five years, after > mining and construction. Almost none of that growth is forecast tocome > from shrinking traditional media.> Dwarf Fortress Economics...hm. > Thank you, social media. (Good article - good comments - angst-ridden > poem (is there really any other kind?)) Republicans, the laggard services still among Moody's et al, yes. (For (poem, other media kind).) From marty at halvorson.us Thu Oct 22 18:18:24 2009 From: marty at halvorson.us (Marty Halvorson) Date: Thu, 22 Oct 2009 19:18:24 -0600 Subject: [FoRK] Redox battery for EVs is recharged in minutes, through electrolyte exchange Message-ID: <4AE10460.7090507@halvorson.us> Stephen Williams wrote: > [This spot for a wild and crazy idea.]: > Create tracking intense light projecting zones on streets, > highways, and especially hills. Use grates, mirrors, > whatever. As photo-rechargeable cars drive over the > charging zones, focused, high intensity light illuminates > just the receptors on the bottom of cars. And all the astronomers, especially the backyard type, will be very pissed off because of the light pollution this would cause. Have you ever tried looking at the sky in the vicinity of movie opening searchlights? It's pretty much impossible to see any stars. Not to belittle the problem of people who drive on muddy roads. Like me these past few days. Peace, Marty Halvorson From jbone at place.org Thu Oct 22 18:47:22 2009 From: jbone at place.org (Jeff Bone) Date: Thu, 22 Oct 2009 20:47:22 -0500 Subject: [FoRK] mappings Message-ID: <8E66CDAC-3FCA-4F2F-8995-688C44BBA494@place.org> re: ConceptNet etc... Stephen quotes me and says: >> Neural networks and other similar systems are something else >> entirely, >> though, and while there's a mapping here it's a bit elusive. >> Spreading activation in semantic networks with fuzzy, defeasible >> semantics seems like a pretty rich topic at present. > > I'm glad you now see a mapping / equivalence. Slow your roll, there, Stephen. I did *not* say there was an equivalence and a mapping is not an equivalence. Neural networks (the perceptron / single-layer / ANN kind) do one thing and one thing only: they statistically learn a discrimination surface in n-feature space. That's all they do and it's all they *can* do, and there are hard computational limits on what that enables. These limits were conclusively demonstrated by Minsky and Papert et. al. in the late 60s (cf. their book Perceptrons) though they both overstated the implications of these limits *and* were largely the genesis, through other people misinterpreting their work, for the almost-complete disappearance of ANN and connectionist-type models from AI and computer science research for a decade and a half. (Which is a damn shame, of course.) More complex neural network wiring schemes have some different properties than e.g. the pure perceptron, but it's still the case that what they do is build classification or discrimination surfaces or predict values according to either a learned linear relationship or (in the case of recurrent networks) some essentially tail-recursive algorithm. What's going on w/ a neural network is *not* semantics; it's formal, it's math, and it's not even particularly topological in itself. Hence the "something else entirely, though." The "while there's a mapping here it's a bit elusive" comment regards the stuff that's being done on the frontier of connectionist research, *not* the traditional ANN but rather its extrapolations in things like hierarchical temporal memories and Geoffrey Hinton's work. I'm *not* stipulating any change of position regarding any earlier debate we had; your insistence that any of these things have any particular *equivalence* is about as useful as saying that hash tables and lists are "equivalent" because they are both examples of data structures. ;-) Just to clarify. jb From dmorton at bitfurnace.com Thu Oct 22 18:53:41 2009 From: dmorton at bitfurnace.com (Damien Morton) Date: Fri, 23 Oct 2009 12:53:41 +1100 Subject: [FoRK] Redox battery for EVs is recharged in minutes through electrolyte exchange In-Reply-To: <4AE020CF.9090603@lig.net> References: <4ADFACC7.7010506@lig.net> <20091022082414.GR27331@leitl.org> <4AE020CF.9090603@lig.net> Message-ID: <8092dc770910221853x1df33bc5m5a8ec484c0f9e900@mail.gmail.com> On Thu, Oct 22, 2009 at 8:07 PM, Stephen Williams wrote: > > [This spot for a wild and crazy idea.]: > Create tracking intense light projecting zones on streets, highways, and > especially hills. Use grates, mirrors, whatever. As photo-rechargeable > cars drive over the charging zones, focused, high intensity light > illuminates just the receptors on the bottom of cars. The intense pulse is > captured by multi-layer / fractal / nano receivers and the impulse is stored > in capacitor banks that slowly drain into the batteries / wheels. Presto: > an optical (or microwave, or even inductive) "third rail". Probable power > scale mismatch, probably safety issues... > > Dude, you just gave me a brilliant idea. Instead of an optical energy transfer system, you could have a simple mechanical energy transfer system. Lets say that in the center of each lane is a slot, and under the slot is a cable moving in the direction the vehicle is travelling. To transfer kinetic energy to the vehicle, the vehicle merely couples itself to the cable. Short term energy storage for travelling off the grid is achieved by a kinetic rolling mass energy system. Advantage of this approach is that it uses no new technology at all, and no fuel or motor is required on the vehicle, enabling mass and cost reductions barely imaginable to the current automotive industry. I estimate that switching to this remarkable new mode of transportation could save anywhere from 10 - 90% of the United State's oil consumption in the transport sector. From sdw at lig.net Thu Oct 22 19:12:10 2009 From: sdw at lig.net (Stephen Williams) Date: Thu, 22 Oct 2009 19:12:10 -0700 Subject: [FoRK] Redox battery for EVs is recharged in minutes, through electrolyte exchange In-Reply-To: <4AE10460.7090507@halvorson.us> References: <4AE10460.7090507@halvorson.us> Message-ID: <4AE110FA.3000707@lig.net> Marty Halvorson wrote: > Stephen Williams wrote: > > [This spot for a wild and crazy idea.]: > > Create tracking intense light projecting zones on streets, > > highways, and especially hills. Use grates, mirrors, > > whatever. As photo-rechargeable cars drive over the > > charging zones, focused, high intensity light illuminates > > just the receptors on the bottom of cars. > > And all the astronomers, especially the backyard type, will be very > pissed off because of the light pollution this would cause. Have you > ever tried looking at the sky in the vicinity of movie opening > searchlights? It's pretty much impossible to see any stars. > > Not to belittle the problem of people who drive on muddy roads. Like > me these past few days. There will be some light pollution if visible light is used, however the phrase "just the receptors on the bottom" was intended to give the idea that spilled light would be minimized. Shine a tight beam up through a grating into a receiver 12 inches from the roadway, black the area around that, black everything in and around the grated area and very little light would make both bounces and escape from under the vehicle. You'd be down to pretty much lateral diffusion from dust. You just turn off the whole system in rain. Snow's less of an issue. There are many areas, like much of California 9 months of the year, where there are no muddy roads at all. In the end, probably not completely feasible or worth doing, but it is at least somewhat feasible. In thinking through the induction possibilities, I'm uncomfortable with that too close to passengers or cabin electronics. Perhaps a panel that extends in front of the vehicle would work, automatically retracting at low speeds. The induction would have to have 3 opposing sections, pos/neg/pos, so that there is no net vertical force. If you combined inductive force, light (maybe daytime only?), and microwave going into the same charging panel, perhaps you could boost the energy delivery quite a bit. If you could get the energy density high enough to make any difference, there are some attractive things about it. A single charging strip could give a boost to thousands of vehicles per day. Strategically placing them could provide enough power to obviate any charging step. Popular stop light lanes, toll plazas, hills, highway entrances. A cheaper trickle version could be used in parking spaces to avoid charging hookups. sdw > > Peace, > Marty Halvorson From jbone at place.org Thu Oct 22 19:25:53 2009 From: jbone at place.org (Jeff Bone) Date: Thu, 22 Oct 2009 21:25:53 -0500 Subject: [FoRK] mappings In-Reply-To: <8E66CDAC-3FCA-4F2F-8995-688C44BBA494@place.org> References: <8E66CDAC-3FCA-4F2F-8995-688C44BBA494@place.org> Message-ID: On Oct 22, 2009, at 8:47 PM, Jeff Bone wrote: > ...your insistence that any of these things have any particular > *equivalence* is about as useful as saying that hash tables and > lists are "equivalent" because they are both examples of data > structures. ;-) (Apologies for the abysmal formatting in the previous; I plead Mail.app. ;-) While the above is pithy and to the point, an even better statement might be that Stephen's conjecture is about as useful as saying that hash tables and lists are "equivalent" because they are both data structures that can be represented by s-expressions (or strings, or whatever.) If I write the following in some Lisp-ish: (('a 1) ('b 2) ('c 3)) ...then, while it has a single representation (given) above, it as several meanings. First, to the reader / tokenizer, it's a string. To the interpreter, it's an s-expression, particularly a nested list s- expr. To some piece of code that implicitly understands the programmer's intent, it's a representation of a hash table or dictionary. THAT'S where the semantics happens. Don't be fooled by the surface similarity of these various models e.g. ANNs vs. Bayesian networks vs. Rete networks vs. semantic networks. You can print them all out as graphs, and at some level working with all of them is just graph traversal. But what the topology and weights of e.g. an ANN *mean* and *do* are quite different from what e.g. a semantic network *means* and *does*. And a strict ontological semantic network is *very* different in its capabilities from a fuzzy, defeasible semantic network. And indeed, even among ANNs there are some very significant differences between what happens with and what can be done with different topologies. The simplest example of this was the genesis of our first conversation about this: a single-layer neural network can have no "statefulness." It can learn, for example, the relationships between all of the inputs it receives simultaneously, and can classify each such set of weights into some set of output classes over any number of input examples; what it *is structurally incapable of doing* is learning the relationships, if any, between *sequences of sets of inputs*. To prove this to yourself, wire up an single-layer / non-recurrent ANN and feed it some input data in some certain order. Then, take the same set-of-sets-of-inputs and feed them to another single-layer non- recurrent ANN *in random order.* Given the same training data, you will end up with identical networks (identical sets of weights) regardless of the order of the sets of inputs fed to it to train it --- the order of *sets of inputs* doesn't matter because the single- layer non-recurrent network has no "memory" or "perception" or "state." It doesn't "remember" the order in which the learn examples occurred; it just learned how to fit an n-line between the various target classes. Point being, this is just one example --- but there's a gestalt lurking here. Don't be too quick to see patterns where their aren't any. Your own internal classifier may not be looking at a large enough volume of the feature space in the potential training set. ;-) $0.02, jb From sdw at lig.net Thu Oct 22 19:37:51 2009 From: sdw at lig.net (Stephen Williams) Date: Thu, 22 Oct 2009 19:37:51 -0700 Subject: [FoRK] Redox battery for EVs is recharged in minutes through electrolyte exchange In-Reply-To: <8092dc770910221853x1df33bc5m5a8ec484c0f9e900@mail.gmail.com> References: <4ADFACC7.7010506@lig.net> <20091022082414.GR27331@leitl.org> <4AE020CF.9090603@lig.net> <8092dc770910221853x1df33bc5m5a8ec484c0f9e900@mail.gmail.com> Message-ID: <4AE116FF.2040708@lig.net> Damien Morton wrote: > On Thu, Oct 22, 2009 at 8:07 PM, Stephen Williams wrote: > > >> [This spot for a wild and crazy idea.]: >> Create tracking intense light projecting zones on streets, highways, and >> especially hills. Use grates, mirrors, whatever. As photo-rechargeable >> cars drive over the charging zones, focused, high intensity light >> illuminates just the receptors on the bottom of cars. The intense pulse is >> captured by multi-layer / fractal / nano receivers and the impulse is stored >> in capacitor banks that slowly drain into the batteries / wheels. Presto: >> an optical (or microwave, or even inductive) "third rail". Probable power >> scale mismatch, probably safety issues... >> >> >> > Dude, you just gave me a brilliant idea. > > Instead of an optical energy transfer system, you could have a simple > mechanical energy transfer system. > I considered mechanical, like something to rev a flywheel or an arm for forward motion. All too crazy. Electric power or burning / expanding fuel is needed. > Lets say that in the center of each lane is a slot, and under the slot is a > cable moving in the direction the vehicle is travelling. To transfer kinetic > energy to the vehicle, the vehicle merely couples itself to the cable. Short > term energy storage for travelling off the grid is achieved by a kinetic > rolling mass energy system. > Lol... The cable car auto revolution we never had. > Advantage of this approach is that it uses no new technology at all, and no > fuel or motor is required on the vehicle, enabling mass and cost reductions > barely imaginable to the current automotive industry. > Installing, maintaining, and powering that cable with its massive friction burden doesn't seem cool. I have pointed out before that we ought to have very efficient train cars with loading as easy as pulling into a parking spot. Drive to a station just off the highway, pull into a slot, have that loaded, potentially in continuous motion on a moving train, then zip along at fairly high speed (100mph+) to the next major node, then drive to your destination. Trains are efficient, but using a train isn't unless you happen to be within very close walking distance on each side. If we could convert individual travel to train-level efficiency, whether using a train or not, that would be a huge win and likely good enough for quite a while. 400 short ton-miles per gallon could translate to 400 miles per gallon per 2000 lb. vehicle, or as little as .25 gallon per 400 miles per person. http://en.wikipedia.org/wiki/Fuel_efficiency_in_transportation#Trains > I estimate that switching to this remarkable new mode of transportation > could save anywhere from 10 - 90% of the United State's oil consumption in > the transport sector. > Lol. You could start with a car that can clip onto existing cable car cables in SF! ;-) Stephen From ken_ganshirt at yahoo.ca Thu Oct 22 19:50:35 2009 From: ken_ganshirt at yahoo.ca (Ken Ganshirt @ Yahoo) Date: Thu, 22 Oct 2009 19:50:35 -0700 (PDT) Subject: [FoRK] Redox battery for EVs is recharged in minutes through electrolyte exchange In-Reply-To: <8092dc770910221853x1df33bc5m5a8ec484c0f9e900@mail.gmail.com> Message-ID: <624385.37437.qm@web33008.mail.mud.yahoo.com> --- On Thu, 10/22/09, Damien Morton wrote: > > I estimate that switching to this remarkable new mode of transportation > could save anywhere from 10 - 90% of the United State's oil > consumption in the transport sector. > The capex alone for the infrastructure build for any significant portion of the continent would be outrageous. Who foots that bill? Then add in the resource cost (the real costs that never get included in the capex) for the infrastructure. You would have to reduce energy consumption for the distances traveled per capita into the negative just to break even. One solution to vehicle energy consumption - speaking of wild ideas - is just stop building cars. Roughly half the energy consumption of any new vehicle is incurred before it rolls a wheel off the dealer's lot. ...ken... __________________________________________________________________ Looking for the perfect gift? Give the gift of Flickr! http://www.flickr.com/gift/ From dmorton at bitfurnace.com Thu Oct 22 20:14:43 2009 From: dmorton at bitfurnace.com (Damien Morton) Date: Fri, 23 Oct 2009 14:14:43 +1100 Subject: [FoRK] Redox battery for EVs is recharged in minutes through electrolyte exchange In-Reply-To: <4AE116FF.2040708@lig.net> References: <4ADFACC7.7010506@lig.net> <20091022082414.GR27331@leitl.org> <4AE020CF.9090603@lig.net> <8092dc770910221853x1df33bc5m5a8ec484c0f9e900@mail.gmail.com> <4AE116FF.2040708@lig.net> Message-ID: <8092dc770910222014i478e5d54m9a97cf8ad102d5db@mail.gmail.com> On Fri, Oct 23, 2009 at 1:37 PM, Stephen Williams wrote: > >> I considered mechanical, like something to rev a flywheel or an arm for > forward motion. All too crazy. Electric power or burning / expanding fuel > is needed. > >> >> A yes, I see it now.... A line of steam powered computer aimed, trebuchets stretching off to the horizon, your vehicle propelled by being launched from one trebuchet to the next. A thousand years hence, people would look back on this transportation system on awe and wonder. From sdw at lig.net Thu Oct 22 20:18:07 2009 From: sdw at lig.net (Stephen Williams) Date: Thu, 22 Oct 2009 20:18:07 -0700 Subject: [FoRK] Redox battery for EVs is recharged in minutes through electrolyte exchange In-Reply-To: <8092dc770910222014i478e5d54m9a97cf8ad102d5db@mail.gmail.com> References: <4ADFACC7.7010506@lig.net> <20091022082414.GR27331@leitl.org> <4AE020CF.9090603@lig.net> <8092dc770910221853x1df33bc5m5a8ec484c0f9e900@mail.gmail.com> <4AE116FF.2040708@lig.net> <8092dc770910222014i478e5d54m9a97cf8ad102d5db@mail.gmail.com> Message-ID: <4AE1206F.3030506@lig.net> Damien Morton wrote: > On Fri, Oct 23, 2009 at 1:37 PM, Stephen Williams wrote: > > >>> I considered mechanical, like something to rev a flywheel or an arm for >>> >> forward motion. All too crazy. Electric power or burning / expanding fuel >> is needed. >> >> >>> > A yes, I see it now.... > > A line of steam powered computer aimed, trebuchets stretching off to the > horizon, your vehicle propelled by being launched from one trebuchet to the > next. > Just solve the acceleration of death problem. A little anti-grav / inertial damper. Wait, now there's a good MOH idea! sdw > A thousand years hence, people would look back on this transportation system > on awe and wonder. > From sdw at lig.net Thu Oct 22 20:19:23 2009 From: sdw at lig.net (Stephen Williams) Date: Thu, 22 Oct 2009 20:19:23 -0700 Subject: [FoRK] Redox battery for EVs is recharged in minutes through electrolyte exchange In-Reply-To: <624385.37437.qm@web33008.mail.mud.yahoo.com> References: <624385.37437.qm@web33008.mail.mud.yahoo.com> Message-ID: <4AE120BB.1010000@lig.net> Ken Ganshirt @ Yahoo wrote: > --- On Thu, 10/22/09, Damien Morton wrote: > > >> I estimate that switching to this remarkable new mode of transportation >> could save anywhere from 10 - 90% of the United State's oil >> consumption in the transport sector. >> >> > > The capex alone for the infrastructure build for any significant portion of the continent would be outrageous. Who foots that bill? > > Then add in the resource cost (the real costs that never get included in the capex) for the infrastructure. You would have to reduce energy consumption for the distances traveled per capita into the negative just to break even. > > One solution to vehicle energy consumption - speaking of wild ideas - is just stop building cars. Roughly half the energy consumption of any new vehicle is incurred before it rolls a wheel off the dealer's lot. > Good point. Or design and build them a lot better. More modular, possible to upgrade, renovate, convert means lasting much longer. Sports car during the week, minivan on weekends, pick up truck, air car for the beach... Automated vehicles can help with a lot of this. The rental shows up just when you need it. Stephen > ...ken... > From sdw at lig.net Thu Oct 22 20:40:10 2009 From: sdw at lig.net (Stephen Williams) Date: Thu, 22 Oct 2009 20:40:10 -0700 Subject: [FoRK] mappings In-Reply-To: <8E66CDAC-3FCA-4F2F-8995-688C44BBA494@place.org> References: <8E66CDAC-3FCA-4F2F-8995-688C44BBA494@place.org> Message-ID: <4AE1259A.5080604@lig.net> Jeff Bone wrote: > > re: ConceptNet etc... > > Stephen quotes me and says: > >>> Neural networks and other similar systems are something else entirely, >>> though, and while there's a mapping here it's a bit elusive. >>> Spreading activation in semantic networks with fuzzy, defeasible >>> semantics seems like a pretty rich topic at present. >> >> I'm glad you now see a mapping / equivalence. > > Slow your roll, there, Stephen. I did *not* say there was an > equivalence and a mapping is not an equivalence. Neural networks (the > perceptron / single-layer / ANN kind) do one thing and one thing > only: they statistically learn a discrimination surface in n-feature > space. That's all they do and it's all they *can* do, and there are > hard computational limits on what that enables. These limits were > conclusively demonstrated by Minsky and Papert et. al. in the late 60s > (cf. their book Perceptrons) though they both overstated the > implications of these limits *and* were largely the genesis, through > other people misinterpreting their work, for the almost-complete > disappearance of ANN and connectionist-type models from AI and > computer science research for a decade and a half. (Which is a damn > shame, of course.) Minsky and Papert were essentially wrong. Sure, they proved all kinds of things about a two layer network. So what? That is like saying you can't build a spaceship with one valve. What they did say was that you couldn't build XOR with 2 layers. How did it not occur to them to add a third layer and additional information flow? > More complex neural network wiring schemes have some different > properties than e.g. the pure perceptron, but it's still the case that > what they do is build classification or discrimination surfaces or > predict values according to either a learned linear relationship or > (in the case of recurrent networks) some essentially tail-recursive > algorithm. > What's going on w/ a neural network is *not* semantics; it's formal, > it's math, and it's not even particularly topological in itself. > Hence the "something else entirely, though." The "while there's a > mapping here it's a bit elusive" comment regards the stuff that's > being done on the frontier of connectionist research, *not* the > traditional ANN but rather its extrapolations in things like > hierarchical temporal memories and Geoffrey Hinton's work. I'm *not* > stipulating any change of position regarding any earlier debate we > had; your insistence that any of these things have any particular > *equivalence* is about as useful as OK. Me either. ;-) I think I made it clear before that I wasn't referring to a particular classical idea of a neural network. I think more in designs that could be called "hierarchical temporal memories", etc. Thinking about the likely way that, in real brains, neurons actually change their structure, connectivity, weighting, and have a very large fanout. That there are competing hypothesis testing and selection going on. Etc. With a rich version of neural nets, I think you start getting close to the kind of structure that is equivalent to the results of automatic training of Markov / Bayesian networks. However, even with a current hidden-layer feedback network, there is some similarity. I didn't mean that they could be mapped completely. You can't even do that completely between Bayesian and Markov probability graphs and they are pretty similar. There are a number of ways to use Bayesian and Markov network ideas to do reasoning, precise and not, and machine learning to train them. Probabilistic graph models, probabilistic logic networks, probabilistic relational reasoning... I'll just call the whole group of concepts / algorithms: Probabilistic Graph Models (PGM) to be generic. It's what Koller used. "Semantics" in a PGM are just probability distributions of possible values of unknown variables along with a reasoning algorithm that prunes the work space as it goes, computing remaining probabilities in any direction given any known variable values. If you have a PGM reasoner output the most probable value of an unknown variable given a certain input, the use is a lot like a NN and the "semantics" could be similar. Maybe the semantics are in training? The simplest (but so inefficient that it is not done) way to train a PGM is to just to assume to start that every variable is dependent on every other variable. This is a full-mesh network to start. You then process a training set, determining for each possible dependency whether there is a valid probability relationship or not. For those that aren't valid over the entire training set, you prune the connection and simplify the model. For a node that unifies a probability effect, you trim direct links and rely on that node. If you have specified intermediate variables required, that should suffice. If not, then you need something like NN hidden layers which either have to be manually specified or auto-discovered. My point was that starting with a raw NN and training it and starting with the equivalent of a full-mesh PGM and training it look the same on the outside. On the inside, current methods for both seem like they are searching for the same goal: high fidelity probabilistic answers to unknowns given partial knowledge. The strategies are completely different, and the degree of structure is different, however it is a similar search. PGMs can embody semantics, including formal, mathematical systems, but with full probability partial knowledge implication. NNs have less form and some interesting characteristics. PGMs already ate expert systems, logic reasoning, and fuzzy logic. I think the distinctiveness of NNs should be added to the PGM collective. And then add in genetic programming (GP not GA). If you start with a hand-picked Bayesian/Markov graph of optimal variables, it doesn't seem anything like a NN. Instead, if you start with a set of variables, assume the equivalent of full-mesh to start, then train by something that approximates computing all possible probability relationships and then trims all dependencies that don't have any strength, it seems a lot closer. Even somewhat dumb hybrids could be interesting. For instance, assume that you have a good PGM training method, but there is no obvious way to find what internal variable nodes you should have. It might be faster to find an algorithm to determine key hidden relationships by training a NN and then analyzing for activation paths to internal nodes from certain variables. You then create these in the PGM and train that. > saying that hash tables and lists are "equivalent" because they are > both examples of data structures. ;-) I don't think that was what I was doing. That was what you thought I was doing. I'm thinking more of the algorithms, use, capability, which all can be similar, and then muse that the representation must have some less than random equivalence distance. sdw > > Just to clarify. > > jb > > _______________________________________________ > FoRK mailing list > http://xent.com/mailman/listinfo/fork From dmorton at bitfurnace.com Fri Oct 23 01:40:46 2009 From: dmorton at bitfurnace.com (Damien Morton) Date: Fri, 23 Oct 2009 19:40:46 +1100 Subject: [FoRK] direct metal laser sintering of Maraging Steel, Titanium, and other metals Message-ID: <8092dc770910230140u75102bdeoa808807a296e9a34@mail.gmail.com> http://www.eos.info/en/products/metal-laser-sintering.html From ken_ganshirt at yahoo.ca Fri Oct 23 09:41:15 2009 From: ken_ganshirt at yahoo.ca (Ken Ganshirt @ Yahoo) Date: Fri, 23 Oct 2009 09:41:15 -0700 (PDT) Subject: [FoRK] direct metal laser sintering of Maraging Steel, Titanium, and other metals In-Reply-To: <8092dc770910230140u75102bdeoa808807a296e9a34@mail.gmail.com> Message-ID: <29636.38901.qm@web33002.mail.mud.yahoo.com> --- On Fri, 10/23/09, Damien Morton wrote: > > http://www.eos.info/en/products/metal-laser-sintering.html > A handy tool to have in the shop when doing restorations. :-) ...ken... __________________________________________________________________ Yahoo! Canada Toolbar: Search from anywhere on the web, and bookmark your favourite sites. Download it now http://ca.toolbar.yahoo.com. From dmorton at bitfurnace.com Fri Oct 23 10:15:47 2009 From: dmorton at bitfurnace.com (Damien Morton) Date: Sat, 24 Oct 2009 04:15:47 +1100 Subject: [FoRK] direct metal laser sintering of Maraging Steel, Titanium, and other metals In-Reply-To: <29636.38901.qm@web33002.mail.mud.yahoo.com> References: <8092dc770910230140u75102bdeoa808807a296e9a34@mail.gmail.com> <29636.38901.qm@web33002.mail.mud.yahoo.com> Message-ID: <8092dc770910231015s5ec507d7nce20d54889011899@mail.gmail.com> Indeed! The only problem with it is that it is too slow. 2-20mm^3/s depending on the metal and the geometry - that's between 50grams and 500grams per hour, and this thing probably has a hourly rate more than a high class hooker. Was reading about one company, AeroMet, that proposed building reactor containment vessels using this technique. Instead of layers of powder being sintered, they would blow an stream of powder onto a surface and laser fuse the stream at the focus point of the laser. With 20 of these laser/metal/aerosol thingies on robot arms/gantries, they estimated they could build a containment vessel in 8 months, complete with cooling cavities etc etc etc. On Sat, Oct 24, 2009 at 3:41 AM, Ken Ganshirt @ Yahoo wrote: > --- On Fri, 10/23/09, Damien Morton wrote: > > > > > http://www.eos.info/en/products/metal-laser-sintering.html > > > > A handy tool to have in the shop when doing restorations. :-) > > ...ken... > > > __________________________________________________________________ > Yahoo! Canada Toolbar: Search from anywhere on the web, and bookmark your > favourite sites. Download it now > http://ca.toolbar.yahoo.com. > _______________________________________________ > FoRK mailing list > http://xent.com/mailman/listinfo/fork > From jbone at place.org Fri Oct 23 11:19:18 2009 From: jbone at place.org (Jeff Bone) Date: Fri, 23 Oct 2009 13:19:18 -0500 Subject: [FoRK] They Laughed at Me When I Said it Message-ID: <7ACF87EF-6FED-4D1F-99C6-EE730C45E0DD@place.org> I can't speak to the credibility of Arnold's source here --- anything involving the word "dominion" in any context freaks me out a little bit --- but this is pretty funny. -- They Laughed at Me When I Said it Arnold Kling http://econlog.econlib.org/archives/2009/10/they_laughed_at.html# The Dominion Post reports [1] The eco-pawprint of a pet dog is twice that of a 4.6-litre Land Cruiser driven 10,000 kilometres a year, researchers have found. Thanks to a reader for the pointer (no pun intended). -- [1] http://www.stuff.co.nz/dominion-post/national/2987821/Save-the-planet-eat-a-dog --- And the comment: "And C4C--Cash for Canines--was born." -- jb From andrew at ceruleansystems.com Fri Oct 23 11:45:41 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Fri, 23 Oct 2009 11:45:41 -0700 Subject: [FoRK] Q re: ConceptNet (also FluidDB) In-Reply-To: <4AE02432.3010906@lig.net> References: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> <4ADF5FE9.6030501@lig.net> <8F8BAAA4-298A-4E49-9C84-42C45EAA52CD@ceruleansystems.com> <4AE02432.3010906@lig.net> Message-ID: <02269ADD-3242-45B1-A4BB-D9C9A29C8853@ceruleansystems.com> On Oct 22, 2009, at 2:21 AM, Stephen Williams wrote: > There was a commercial database product with a very nice Java-based > GUI that was essentially an RDF-graph based database. It was not > hard to understand or use. Unfortunately, they took the wrong > product path. I had several conversations with the founder. He > just deviate from his particular viewpoint and he didn't know > anything about RDF / semantic ideas. > > Anyway, yes, that is true now, however I don't think it is > necessarily true. We just need the equivalent of a spreadsheet > application to make it accessible. I'm thinking about it, when I > can... Product design is a secondary problem. For real-world semantic/RDF graphs, the operating rule of thumb is that current technologies become useless at around 100k nodes for lack of scalability. That is single-workstation in-memory scale. There are a number of companies offering semantic graph engines and customers insensitive to interface with large buckets of cash for a decent graph database but no one is crossing that chasm. >> I've long wondered if there is a relationship between Braess' >> Paradox and pruning algorithms for graph-like computational models >> but am too busy/lazy to spend much time thinking about it. > > Seems like a different thing to me... Although I wouldn't be > surprised if there is some analog. You almost never see these two things together. In the few cases where I have, I always have the nagging feeling that I am looking at separate manifestations of some common principle we haven't figured out. Like I said, idle speculation. From ken_ganshirt at yahoo.ca Fri Oct 23 12:15:10 2009 From: ken_ganshirt at yahoo.ca (Ken Ganshirt @ Yahoo) Date: Fri, 23 Oct 2009 12:15:10 -0700 (PDT) Subject: [FoRK] They Laughed at Me When I Said it In-Reply-To: <7ACF87EF-6FED-4D1F-99C6-EE730C45E0DD@place.org> Message-ID: <320246.10748.qm@web33003.mail.mud.yahoo.com> --- On Fri, 10/23/09, Jeff Bone wrote: > > I can't speak to the credibility of Arnold's source here > --- anything involving the word "dominion" in any context > freaks me out a little bit ... > In this context (the name of a newspaper) it's merely a synomym for any place that was within the domain of the old British Empire. It's the Americans who are exhibiting the imperialist tendencies these days. Perhaps even more freaky, at least to those of us who live right next door and have lots of oil.... ...ken... __________________________________________________________________ Looking for the perfect gift? Give the gift of Flickr! http://www.flickr.com/gift/ From jbone at place.org Fri Oct 23 13:15:35 2009 From: jbone at place.org (Jeff Bone) Date: Fri, 23 Oct 2009 15:15:35 -0500 Subject: [FoRK] mappings Message-ID: Further clarification for Stephen... > With a rich version of neural nets, I think you start getting close > to the kind of structure that is equivalent to the results of > automatic training of Markov / Bayesian networks So to some extent we're talking past each other, largely because the terminological rigor in the field in general has become so sloppy that anything vaguely connectionist-looking is called a "neural network." So fair enough on that point. Back to a couple of your specific examples: there are even substantial differences between ANNs and e.g. Bayesian networks. The former, as mentioned, can *only* learn a classification n-line and, unless recurrent, can learn "moment in time" snapshot-patterns but not patterns about things that develop over time. (Don't get hung up about the recurrency issue; even recurrent ANNs with the typical model and learning algorithms aren't that powerful. I'm not talking about the XOR problem, and no, Minsky and Papert weren't "discredited" in this; there analysis was sound, it was just over-interpreted by them and everyone else.) What an ANN encodes, after the learning phase (in the case of supervised learning, the most common case for the real-world use of these things) is really nothing more than a classifying function: how to fit an n-line to divide the world and make an either-or decision about the inputs with respect to each output. *At best* you could say that the semantic value of the weightings learned encodes some opaque and formal model of the target space with respect to the learned input examples. It's *at best* correlative. The Bayesian networks do something slightly different. They are *explicitly* probabilistically causal with respect to the conditional dependencies between their entity and *explicitly* probabilistic with respect what they're learning; what they're learning is, in effect, an n-dimensional probability density function relating inputs to outputs. This is *far* more general than e.g. the traditional ANN (regardless of the ANN's topology.) If you like, you can give it a similar geometric interpretation: let's say the data is described by 3 features / dimensions; if so, then the ANN learns to fit a plane between examples in 3-space to classify the data. *Roughly* analogously, what the Bayesian network learns to do is more abstract and powerful: it learns to build a kind of "fuzzy field" which can be used to split the feature space --- and this can be said with certainty to be a more robust "model" of the observations than e.g. the model encoded in an ANN's networks. More powerful math, .: more powerful model. Your further example in your equivalenc^H^H^H^H "similarity" class is Markov models. Markov models can be understood as a weaker variant encoding of the kinds of conditional dependencies that you see in a full Bayesian network. There's *actually* a pretty good correspondence between the two, though the Bayesian networks and learning algorithms over them are more abstract. (Consider whether a Bayesian network can "learn" a Markov model. Then consider the converse. Consider in the context of generalization over unseen data.) That said, it's all about what you're actually attempting to do. Surprisingly many real-world phenomenon can be well-understood (at least to the level of decent prediction) without even directly modeling any sort of conditional dependency, logical entailment, etc. HTMs are just a biologically-inspired, turbocharged BN at some level. They employ an online-learning algorithm, some more complex and layered topology, and --- critically --- an a priori semantics imposed implicitly by the learning method, one which considers spatial and temporal aspects of its inputs per se. That's a reasonable thing to do from a biological metaphor perspective: it's like differentiating inputs and regions of the network based on which sense is providing the data, which of course the human neocortex and support systems *do.* But that's a far cry from what you see in the usual ANN or BN, so I would say that an HTM is a *highly advanced* form of BN, almost to the point of no longer really being a BN. (You certainly could use an HTM where a BN would work, but why would you want to? But there are many things for which an HTM may be suited that would be entirely unsuitable applications for a BN.) Small nit: > PGMs can embody semantics, including formal, mathematical systems, > but with full probability partial knowledge implication. If you've got semantics, then your system isn't "formal." Definitionally. ;-) (But yes, I understand what you're attempting to say here.) jb From sdw at lig.net Fri Oct 23 13:58:20 2009 From: sdw at lig.net (Stephen D. Williams) Date: Fri, 23 Oct 2009 13:58:20 -0700 Subject: [FoRK] They Laughed at Me When I Said it In-Reply-To: <7ACF87EF-6FED-4D1F-99C6-EE730C45E0DD@place.org> References: <7ACF87EF-6FED-4D1F-99C6-EE730C45E0DD@place.org> Message-ID: <4AE218EC.1020502@lig.net> Jeff Bone wrote: > ... > They Laughed at Me When I Said it > Arnold Kling > > http://econlog.econlib.org/archives/2009/10/they_laughed_at.html# > The Dominion Post reports [1] > > The eco-pawprint of a pet dog is twice that of a 4.6-litre Land > Cruiser driven 10,000 kilometres a year, researchers have found. Many people have dogs rather than a human, that's the real alternative. And how do you think that footprint compares? sdw From sdw at lig.net Fri Oct 23 14:13:16 2009 From: sdw at lig.net (Stephen D. Williams) Date: Fri, 23 Oct 2009 14:13:16 -0700 Subject: [FoRK] Q re: ConceptNet (also FluidDB) In-Reply-To: <02269ADD-3242-45B1-A4BB-D9C9A29C8853@ceruleansystems.com> References: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> <4ADF5FE9.6030501@lig.net> <8F8BAAA4-298A-4E49-9C84-42C45EAA52CD@ceruleansystems.com> <4AE02432.3010906@lig.net> <02269ADD-3242-45B1-A4BB-D9C9A29C8853@ceruleansystems.com> Message-ID: <4AE21C6C.6000407@lig.net> J. Andrew Rogers wrote: > On Oct 22, 2009, at 2:21 AM, Stephen Williams wrote: >> There was a commercial database product with a very nice Java-based >> GUI that was essentially an RDF-graph based database. It was not >> hard to understand or use. Unfortunately, they took the wrong >> product path. I had several conversations with the founder. He just >> deviate from his particular viewpoint and he didn't know anything >> about RDF / semantic ideas. >> >> Anyway, yes, that is true now, however I don't think it is >> necessarily true. We just need the equivalent of a spreadsheet >> application to make it accessible. I'm thinking about it, when I can... > > Product design is a secondary problem. That is only true for certain classes of problems. > > For real-world semantic/RDF graphs, the operating rule of thumb is > that current technologies become useless at around 100k nodes for lack > of scalability. That is single-workstation in-memory scale. There are > a number of companies offering semantic graph engines and customers > insensitive to interface with large buckets of cash for a decent graph > database but no one is crossing that chasm. > Scalability is an issue. On the other hand, most scalability issues have a solution. Certainly simple, flat triple stores aren't going to do it. I introduced chunkiness to one of my designs (it had temporal versioning.) Other ideas include certain kinds of clustering, denormalization-like constructs, etc. How would you characterize the scalability problems that you have seen? What fundamental issue was involved? sdw From geege4 at gmail.com Fri Oct 23 14:27:31 2009 From: geege4 at gmail.com (geege schuman) Date: Fri, 23 Oct 2009 17:27:31 -0400 Subject: [FoRK] Clever, Indeed Message-ID: <493a95a00910231427g3b3923d5x6eb45b262dd6b0b6@mail.gmail.com> http://adage.com/article?article_id=139862 Geege From andrew at ceruleansystems.com Fri Oct 23 14:46:00 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Fri, 23 Oct 2009 14:46:00 -0700 Subject: [FoRK] Q re: ConceptNet (also FluidDB) In-Reply-To: <4AE21C6C.6000407@lig.net> References: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> <4ADF5FE9.6030501@lig.net> <8F8BAAA4-298A-4E49-9C84-42C45EAA52CD@ceruleansystems.com> <4AE02432.3010906@lig.net> <02269ADD-3242-45B1-A4BB-D9C9A29C8853@ceruleansystems.com> <4AE21C6C.6000407@lig.net> Message-ID: <1435D16A-45C0-486E-8682-05907D69ADA7@ceruleansystems.com> On Oct 23, 2009, at 2:13 PM, Stephen D. Williams wrote: > Scalability is an issue. On the other hand, most scalability issues > have a solution. Certainly simple, flat triple stores aren't going > to do it. I introduced chunkiness to one of my designs (it had > temporal versioning.) Other ideas include certain kinds of > clustering, denormalization-like constructs, etc. > > How would you characterize the scalability problems that you have > seen? What fundamental issue was involved? To be clear, I've never used the various triple stores (of which there are myriad designs) out there. I do work with people for whom it is an important problem. The fundamental issue is dynamic analytic performance at non-trivial scales. One could say something similar about all databases, for very similar technical reasons, but the limitations manifest much earlier in graph databases. From dmorton at bitfurnace.com Fri Oct 23 15:15:14 2009 From: dmorton at bitfurnace.com (Damien Morton) Date: Sat, 24 Oct 2009 09:15:14 +1100 Subject: [FoRK] Q re: ConceptNet (also FluidDB) In-Reply-To: <1435D16A-45C0-486E-8682-05907D69ADA7@ceruleansystems.com> References: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> <4ADF5FE9.6030501@lig.net> <8F8BAAA4-298A-4E49-9C84-42C45EAA52CD@ceruleansystems.com> <4AE02432.3010906@lig.net> <02269ADD-3242-45B1-A4BB-D9C9A29C8853@ceruleansystems.com> <4AE21C6C.6000407@lig.net> <1435D16A-45C0-486E-8682-05907D69ADA7@ceruleansystems.com> Message-ID: <8092dc770910231515r4cf79e64w13215cd6529658f1@mail.gmail.com> The only way a triple store can work is to use the triple-store as a conceptual model while building out indices and denormalisations based on the queries applied to the store. There was an in-memory C++ data store from a dude in russia that did something like that - the queries needed to be all provided at compile time and a data-store suitable for those queries was created as part of the compilation process. On Sat, Oct 24, 2009 at 8:46 AM, J. Andrew Rogers < andrew at ceruleansystems.com> wrote: > > On Oct 23, 2009, at 2:13 PM, Stephen D. Williams wrote: > >> Scalability is an issue. On the other hand, most scalability issues have >> a solution. Certainly simple, flat triple stores aren't going to do it. I >> introduced chunkiness to one of my designs (it had temporal versioning.) >> Other ideas include certain kinds of clustering, denormalization-like >> constructs, etc. >> >> How would you characterize the scalability problems that you have seen? >> What fundamental issue was involved? >> > > > To be clear, I've never used the various triple stores (of which there are > myriad designs) out there. I do work with people for whom it is an important > problem. > > The fundamental issue is dynamic analytic performance at non-trivial > scales. One could say something similar about all databases, for very > similar technical reasons, but the limitations manifest much earlier in > graph databases. > > _______________________________________________ > FoRK mailing list > http://xent.com/mailman/listinfo/fork > From beberg at mithral.com Fri Oct 23 15:42:46 2009 From: beberg at mithral.com (Adam L Beberg) Date: Fri, 23 Oct 2009 15:42:46 -0700 Subject: [FoRK] They Laughed at Me When I Said it In-Reply-To: <7ACF87EF-6FED-4D1F-99C6-EE730C45E0DD@place.org> References: <7ACF87EF-6FED-4D1F-99C6-EE730C45E0DD@place.org> Message-ID: <4AE23166.7040202@mithral.com> Jeff Bone wrote on 10/23/2009 11:19 AM: > The eco-pawprint of a pet dog is twice that of a 4.6-litre Land Cruiser > driven 10,000 kilometres a year, researchers have found. Ahh, that's the thing isn't it. Didn't the energy your dog uses to run around come from under the ground to be spewed into the air, or was it there already. A wild wolf is carbon and energy neutral. So is a aboriginal human living in the Amazon. But your atoms and bits are probably all dug up. Coal power plants, fuel and chemicals dumped on the food and diesel to transport it. Your Pet's food, house, and collar are all oil based too. So the article is wrong, the proper phrasing is... Save the planet, die and seal the carbon deep underground. We're way way past the tipping point. Unless we TOMORROW stop all fossil fuel use and humans do nothing but work 24/7 for the next 100 years sequestering carbon back into the ground, this planet is completely dead as far as us oxygen<->CO2 cycle lifeforms are concerned. Maybe that wouldn't even help. Someone better get their ass in gear on establishing an off planet colony soon, before the wars knock the tech level back down below what we need to get off planet. -- Adam L. Beberg http://www.mithral.com/~beberg/ From sdw at lig.net Fri Oct 23 15:48:23 2009 From: sdw at lig.net (Stephen D. Williams) Date: Fri, 23 Oct 2009 15:48:23 -0700 Subject: [FoRK] Q re: ConceptNet (also FluidDB) In-Reply-To: <8092dc770910231515r4cf79e64w13215cd6529658f1@mail.gmail.com> References: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> <4ADF5FE9.6030501@lig.net> <8F8BAAA4-298A-4E49-9C84-42C45EAA52CD@ceruleansystems.com> <4AE02432.3010906@lig.net> <02269ADD-3242-45B1-A4BB-D9C9A29C8853@ceruleansystems.com> <4AE21C6C.6000407@lig.net> <1435D16A-45C0-486E-8682-05907D69ADA7@ceruleansystems.com> <8092dc770910231515r4cf79e64w13215cd6529658f1@mail.gmail.com> Message-ID: <4AE232B7.9070309@lig.net> Damien Morton wrote: > The only way a triple store can work is to use the triple-store as a > conceptual model while building out indices and denormalisations based on > the queries applied to the store. > There was an in-memory C++ data store from a dude in russia that did > something like that - the queries needed to be all provided at compile time > and a data-store suitable for those queries was created as part of the > compilation process. > That's part of my design for a solution, except at run time, not compile time. That is a huge part of the solution. It works as if it is an idealized triple store, but can choose between multiple internal representations depending on optimization. Along with a binary RDF interchange format, inspired by my W3C EXI participation but with many additional ideas. This is the kind of thing that would be good to do as an open source project. Too hard to get something commercial to revenue producing stage, but needed for many things. The hard part seems more at the query / inference level, not optimizing storage. Perhaps just because there is more complexity and I haven't tried to solve it yet. Note though that you can get pretty far with a straight triple store by observing a few things. First, the actual triples only need to be sets of 3 integers. All strings are in a separate string table / inverted index / regexp scanable store. They are stripped/mapped on input and restored on output, which could even be done on a separate tier / cloud or at the client. If you use variable integer format, that means 6-9 bytes per triple typically in main memory. The single "table" of triples then gets indexed 6 times for each pair ordering of the triple elements pointing to the third. This is the expensive part of writes, especially if you also do any clustering and stats maintenance. After that base, the fun starts. I'm a year or two out of date of current techniques, however it seemed like few projects were getting very creative on the scalability side. There are some nice commercial products, although most of what I see is a lot of advancement on the query / reasoning side. Stephen > On Sat, Oct 24, 2009 at 8:46 AM, J. Andrew Rogers < > andrew at ceruleansystems.com> wrote: > > >> On Oct 23, 2009, at 2:13 PM, Stephen D. Williams wrote: >> >> >>> Scalability is an issue. On the other hand, most scalability issues have >>> a solution. Certainly simple, flat triple stores aren't going to do it. I >>> introduced chunkiness to one of my designs (it had temporal versioning.) >>> Other ideas include certain kinds of clustering, denormalization-like >>> constructs, etc. >>> >>> How would you characterize the scalability problems that you have seen? >>> What fundamental issue was involved? >>> >>> >> To be clear, I've never used the various triple stores (of which there are >> myriad designs) out there. I do work with people for whom it is an important >> problem. >> >> The fundamental issue is dynamic analytic performance at non-trivial >> scales. One could say something similar about all databases, for very >> similar technical reasons, but the limitations manifest much earlier in >> graph databases. >> >> _______________________________________________ >> FoRK mailing list >> http://xent.com/mailman/listinfo/fork >> >> > _______________________________________________ > FoRK mailing list > http://xent.com/mailman/listinfo/fork > From andrew at ceruleansystems.com Fri Oct 23 15:53:39 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Fri, 23 Oct 2009 15:53:39 -0700 Subject: [FoRK] Q re: ConceptNet (also FluidDB) In-Reply-To: <8092dc770910231515r4cf79e64w13215cd6529658f1@mail.gmail.com> References: <2ED1DCC5-5040-4CD4-B113-8071207EBA09@place.org> <4ADF5FE9.6030501@lig.net> <8F8BAAA4-298A-4E49-9C84-42C45EAA52CD@ceruleansystems.com> <4AE02432.3010906@lig.net> <02269ADD-3242-45B1-A4BB-D9C9A29C8853@ceruleansystems.com> <4AE21C6C.6000407@lig.net> <1435D16A-45C0-486E-8682-05907D69ADA7@ceruleansystems.com> <8092dc770910231515r4cf79e64w13215cd6529658f1@mail.gmail.com> Message-ID: <4D27776E-70EE-46EA-8E98-285F714E27D9@ceruleansystems.com> On Oct 23, 2009, at 3:15 PM, Damien Morton wrote: > The only way a triple store can work is to use the triple-store as a > conceptual model while building out indices and denormalisations > based on > the queries applied to the store. > There was an in-memory C++ data store from a dude in russia that did > something like that - the queries needed to be all provided at > compile time > and a data-store suitable for those queries was created as part of the > compilation process. Yep. A dozen variations on this basic idea is how it usually works today. There are a few people playing with literal graphs under the hood, but that requires exotic hardware most people do not have to work adequately. From sdw at lig.net Fri Oct 23 15:54:51 2009 From: sdw at lig.net (Stephen D. Williams) Date: Fri, 23 Oct 2009 15:54:51 -0700 Subject: [FoRK] They Laughed at Me When I Said it In-Reply-To: <4AE23166.7040202@mithral.com> References: <7ACF87EF-6FED-4D1F-99C6-EE730C45E0DD@place.org> <4AE23166.7040202@mithral.com> Message-ID: <4AE2343B.4020808@lig.net> Adam L Beberg wrote: > ... > > Someone better get their ass in gear on establishing an off planet > colony soon, before the wars knock the tech level back down below what > we need to get off planet. Yea, go for it. Whatever you gotta believe to get it done. We really need a religious angle to this too. NASA finds 72,000 virgins on a desert planet in the next star system over? Jesus helps those who helps themselves [get off planet]? Mayans predicted the mega-asteroid to end their calendar? Stephen From jbone at place.org Sat Oct 24 09:26:21 2009 From: jbone at place.org (Jeff Bone) Date: Sat, 24 Oct 2009 11:26:21 -0500 Subject: [FoRK] Drudge vs. USD Message-ID: This is *awesome.* I would *love* to see more folks doing more of this (very simple) type of analysis! http://paul.kedrosky.com/archives/2009/10/matt_drudge_vs.html I call this sort of thing --- the kind of thing that you can do in just a few minutes with R and some other tools, if you've got usable data --- "scratch" analysis. It's a lot like back-of-the-envelope engineering or Fermi estimation; not something to be considered comprehensive / definitive / authoritative, but enough to draw some tentative conclusions. (FWIW, about a half-hour of futzing with R and some publicly-available time series data enabled me to *finally* convince some climate-change denier family members that *at least* there was a real warming trend and that there was *at least* a significant correlation between industrialization and this warming trend.) (One of the projects I've long pondered but will probably never get to is doing some kind of automatic conversational network and topic model (LSA plus a few other "views" of the data) over the history of FoRK and then attempting some scratch analysis on it. The first part's enough of a "research" topic that I'll probably never get over the hump on it, pretty low on my priority list; would be interesting, though.) More like this! jb From jbone at place.org Sat Oct 24 09:29:40 2009 From: jbone at place.org (Jeff Bone) Date: Sat, 24 Oct 2009 11:29:40 -0500 Subject: [FoRK] Existential Risks via Hanson: Bad News on Human Extinction Message-ID: <8441A2C4-D13F-4BAA-93F1-A3D6B86C599F@place.org> From Robin via Overcoming Bias: http://www.overcomingbias.com/2009/10/bad-news-on-human-extinction.html Bad News On Human Extinction By Robin Hanson ? October 23, 2009 2:15 pm ? Discuss ? ? Prev ? Next ? Disasters that destroy all but a thousand humans are more likely [1] than disasters that destroy all but a hundred humans. So this news [2] says human extinction is more likely than we thought: Conservation biologists may be deluding themselves. An analysis of the minimum number of individuals needed for a species to survive in the long term has found that current conservation practices underestimate the risk of extinction by not fully allowing for the dangers posed by the loss of genetic diversity. If correct, it means the number of individuals in endangered species are being allowed to dwindle too far. Lochran Traill at the University of Adelaide, Australia, and colleagues found that for thousands of species the minimum viable population size (MVP) ? where a species has a 90 per cent chance of surviving the next 100 years ? comes in at thousands rather than hundreds of individuals. Many biologists, Traill says, work with lower numbers and so allow unacceptably high extinction risks. [1] http://hanson.gmu.edu/collapse.pdf [2] http://www.newscientist.com/article/dn17974-conservation-targets-too-low-to-save-atrisk-species.html -- jb From jbone at place.org Sat Oct 24 09:54:48 2009 From: jbone at place.org (Jeff Bone) Date: Sat, 24 Oct 2009 11:54:48 -0500 Subject: [FoRK] National Defense is Not a Public Good (proof; interesting, useless) Message-ID: <684C36BC-7F51-4F3C-B5DF-929D0974F3DF@place.org> Nice proof, but relatively useless given reality and game theory. (I.e., it's only *not* a public good if nobody else has it. And with any possibility of eventual "defection" from universal non- malevolence, you're screwed.) http://econlog.econlib.org/archives/2009/10/a_simple_proof.html jb From sdw at lig.net Sat Oct 24 17:45:13 2009 From: sdw at lig.net (Stephen Williams) Date: Sat, 24 Oct 2009 17:45:13 -0700 Subject: [FoRK] mappings In-Reply-To: References: Message-ID: <4AE39F99.1000301@lig.net> Jeff Bone wrote: > > Further clarification for Stephen... > >> With a rich version of neural nets, I think you start getting close >> to the kind of structure that is equivalent to the results of >> automatic training of Markov / Bayesian networks > > So to some extent we're talking past each other, largely because the > terminological rigor in the field in general has become so sloppy that > anything vaguely connectionist-looking is called a "neural network." > So fair enough on that point. Agreed, a friend who has spent 9+ years working on AI/AGI doesn't consider her connectionist work to be a NN at all... I think that all kinds of structures arise in real neural networks that we can't detect well yet, so the fact that our first imitations were primitive and overly simple doesn't obviate the ongoing use of the term to me. I suppose I would use "connectionist" when doing something radically different than what a NN could do. Anyway, connectionist is a reasonable general term for the whole area. > > Back to a couple of your specific examples: there are even > substantial differences between ANNs and e.g. Bayesian networks. The > former, as mentioned, can *only* learn a classification n-line and, > unless recurrent, can learn "moment in time" snapshot-patterns but not > patterns about things that develop over time. (Don't get hung up > about the recurrency issue; even recurrent ANNs with the typical > model and learning algorithms aren't that powerful. I'm not talking > about the XOR problem, and no, Minsky and Papert weren't "discredited" > in this; there analysis was sound, it was just over-interpreted by > them and everyone else.) Which means that their interpretation was discredited. Same gist. > ... > Your further example in your equivalenc^H^H^H^H "similarity" class is > Markov models. Markov models can be understood as a weaker variant > encoding of the kinds of conditional dependencies that you see in a > full Bayesian network. There's *actually* a pretty good > correspondence between the two, though the Bayesian networks and > learning algorithms over them are more abstract. (Consider whether a > Bayesian network can "learn" a Markov model. Then consider the > converse. Consider in the context of generalization over unseen data.) A Bayesian model can be converted to a Markov model, and vice versa. The graph changes in certain cases, but the resulting capability is the same, or the same in most cases. Just the first paper found to support this: http://www.cs.ru.nl/~peterl/markoveq.pdf > > That said, it's all about what you're actually attempting to do. > Surprisingly many real-world phenomenon can be well-understood (at > least to the level of decent prediction) without even directly > modeling any sort of conditional dependency, logical entailment, etc. > > HTMs are just a biologically-inspired, turbocharged BN at some level. > They employ an online-learning algorithm, some more complex and > layered topology, and --- critically --- an a priori semantics imposed > implicitly by the learning method, one which considers spatial and > temporal aspects of its inputs per se. That's a reasonable thing to > do from a biological metaphor perspective: it's like differentiating > inputs and regions of the network based on which sense is providing > the data, which of course the human neocortex and support systems > *do.* But that's a far cry from what you see in the usual ANN or BN, > so I would say that an HTM is a *highly advanced* form of BN, almost > to the point of no longer really being a BN. (You certainly could use > an HTM where a BN would work, but why would you want to? But there > are many things for which an HTM may be suited that would be entirely > unsuitable applications for a BN.) > > Small nit: > >> PGMs can embody semantics, including formal, mathematical systems, >> but with full probability partial knowledge implication. > > If you've got semantics, then your system isn't "formal." > Definitionally. ;-) > > (But yes, I understand what you're attempting to say here.) http://en.wikipedia.org/wiki/Semantics > Semantics[1] is the study of meaning, usually in language. I know that "semantics" is used in several ways, however this seems consistent with "semantic technology" a la RDF a la representation of knowledge rather than disconnected data. A PGM explicitly encodes: "given a, b, and c, then we know d value 1 is x% likely, value 2 is y% likely, ...". http://en.wikipedia.org/wiki/Formal_system > ... a formal system ... consists of a formal language together with a deductive system ... which consists of a set of inference rules and/or axioms A Bayesian/Markov PGM seems to also fit that as any logical inference, including xor, can be represented by the conditional probability models. The deductive system is the reasoning algorithm that computes the remaining conditional probabilities given certain knowledge. A PGM is usually not used as a formal logic system in the full sense, although it should be capable of that. But did you mean something more fundamental? Formal systems can't have or map to meaning? sdw > > > jb > > > > > _______________________________________________ > FoRK mailing list > http://xent.com/mailman/listinfo/fork From jbone at place.org Sun Oct 25 04:54:13 2009 From: jbone at place.org (Jeff Bone) Date: Sun, 25 Oct 2009 06:54:13 -0500 Subject: [FoRK] Cutting through the stimulus BS Message-ID: <8E4B57F8-639B-46BE-8EA2-B9AB98437D0C@place.org> (Via Kling.) http://www.stanford.edu/~johntayl/Evidence%20on%20the%20Sources%20of%20Improved%20Economic%20Growth%20from%20the%20First%20to%20Second%20Quarter.pdf So -- the "improvement" we've seen in GDP growth (mind you, not adjusted for inflationary / dilutionary currency effects, offsetting deflationary cost reductions, etc.) is *primarily* due to increased private investment (in equipment, as it turns out; refresh those tools, boys!) The contribution of any "stimulus" so far is de minimus relative to that. Hardly surprising. Well, except perhaps to a naif or a neo- Keynesian. jb From sdw at lig.net Sun Oct 25 13:18:28 2009 From: sdw at lig.net (Stephen Williams) Date: Sun, 25 Oct 2009 13:18:28 -0700 Subject: [FoRK] Cutting through the stimulus BS In-Reply-To: <8E4B57F8-639B-46BE-8EA2-B9AB98437D0C@place.org> References: <8E4B57F8-639B-46BE-8EA2-B9AB98437D0C@place.org> Message-ID: <4AE4B294.90509@lig.net> Jeff Bone wrote: > > (Via Kling.) > > > http://www.stanford.edu/~johntayl/Evidence%20on%20the%20Sources%20of%20Improved%20Economic%20Growth%20from%20the%20First%20to%20Second%20Quarter.pdf > > > So -- the "improvement" we've seen in GDP growth (mind you, not > adjusted for inflationary / dilutionary currency effects, offsetting > deflationary cost reductions, etc.) is *primarily* due to increased > private investment (in equipment, as it turns out; refresh those > tools, boys!) > > The contribution of any "stimulus" so far is de minimus relative to > that. Hardly surprising. Well, except perhaps to a naif or a > neo-Keynesian. It's all about psychological effects isn't it? These are often indirect. The optimism of doing something via the stimulus somewhat offset the cries of doom, gloom, and disaster long enough for people to take a breather and get back to living and working more or less normally. (Except the unemployed, who probably have already passed through the angst of financial disaster and have found a fall-back position.) The thing that I don't like about the stimulus choices is that they gave away too much money with too little direct positive benefit. The indirect benefits are probably much larger, so it is possible they were good choices since anything more thoughtful up front might have been too slow and too fraught with specific idea backlash to get the immediate effect needed. For instance, I would have favored spending a whole lot to bootstrap a new auto industry, more or less as a clean break from the existing one. Set up design / R&D competitions / markets in the major knowledge-worker areas (Bay Area, DC, Boston, Research Triangle, maybe Austin, etc.), plus open to virtual teams / companies, and then target a few very inexpensive areas for POC manufacturing, perhaps in Wichita (Cessna, etc.) or similar. Near major entertainment areas perhaps, Vegas, Colorado, some nice areas in Canada. Make a market where all manufacturing can attempt to compete. Start serious efforts to create inexpensive "company town" like environments, except not actually owned by the companies involved. Perhaps something between a homeowners association and a credit-union like structure. A key set of ideas would be to make this a push for highly capable robotic factories, cheap, modular vehicles, electric or other low-maintenance hyper efficient propulsion, and AI automated self-driving (a la Darpa Challenge) vehicles. Resolve to solve the technical, economic, and legal issues. After success, plan to solve the efficient auto / train integration and automated aircars. We have plenty of technology to get started on all of that, probably having success in 5 years. We just have to collectively decide to do it. sdw > > jb From wulfconsumer at gmail.com Sun Oct 25 17:11:37 2009 From: wulfconsumer at gmail.com (Wulf Losee) Date: Sun, 25 Oct 2009 17:11:37 -0700 Subject: [FoRK] They Laughed at Me When I Said it Message-ID: <14d68d3f0910251711m6a26a248yee7b32cd6dbdf388@mail.gmail.com> Adam L Beberg wrote: >We're way way past the tipping point. Unless we TOMORROW stop all fossil >fuel use and humans do nothing but work 24/7 for the next 100 years >sequestering carbon back into the ground, this planet is completely dead >as far as us oxygen<->CO2 cycle lifeforms are concerned. Maybe that >wouldn't even help. It may be the end of civilization, but it's won't be the end of the world. Over most of geologic time, since life started modifying our atmosphere, our planet was hotter and more humid than we know it to be today -- with more CO2 in the atmosphere than we have today -- and higher CO2 levels than most climate models predict will be spewed out in the next century (And we'll run out of petroleum and coal in the next 30 years, so I don't think it will go much farther). The notable climate exception to the hotter/warmer Earth was 300 million years ago during the late Carboniferous Period -- which is the only period in the geologic history that resembles our current climate. Rather than running in circles and crying that the sky is falling, we could start sequestering Carbon by dumping iron oxide into the nutrient dead-zones of the the oceans to create plankton blooms. I keep hearing arguments that this won't work. Well, let's get of our collective asses, and see if it works. Also, some folks at Lawrence Berkeley argue that this won't work, because some large percentage of the dead plankton doesn't actually reach the sea-floor. Well, my answer to that, is as long as the Carbon is in dead plankton, the Carbon isn't in the atmosphere -- so it doesn't matter if it's on the sea floor. And if you really really need to get a higher percentage of Carbon sequestered on sea floor, just dump more iron oxide. Also, you wrote: >A wild wolf is carbon and energy neutral. So is a aboriginal human >living in the Amazon. Well, aboriginal humans are far from Carbon neutral. Amazonian aborigines practice slash and burn agriculture, which actually releases a whole lot of CO2. Likewise, hunter gatherers are known to light forest fires to cause game to flee from cover. Pre-Industrial man may have had less of a Carbon footprint than we have today, but it was by no means insignificant. best regards, --Wulf From andrew at ceruleansystems.com Sun Oct 25 17:42:17 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Sun, 25 Oct 2009 17:42:17 -0700 Subject: [FoRK] They Laughed at Me When I Said it In-Reply-To: <14d68d3f0910251711m6a26a248yee7b32cd6dbdf388@mail.gmail.com> References: <14d68d3f0910251711m6a26a248yee7b32cd6dbdf388@mail.gmail.com> Message-ID: On Oct 25, 2009, at 5:11 PM, Wulf Losee wrote: > (And we'll run out of petroleum and coal in the next 30 years, so I > don't think it will go > much farther). Even the most conservative projections (proven recoverable supplies only and assuming exponentially growing consumption rate) put current coal supplies at the better part of a century. We won't be running out for the foreseeable future. Whether or not we run out of petroleum in 30 years is more of a political question than one of theoretical supply. From andrew at ceruleansystems.com Sun Oct 25 18:39:17 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Sun, 25 Oct 2009 18:39:17 -0700 Subject: [FoRK] Cutting through the stimulus BS In-Reply-To: <4AE4B294.90509@lig.net> References: <8E4B57F8-639B-46BE-8EA2-B9AB98437D0C@place.org> <4AE4B294.90509@lig.net> Message-ID: <95CB8FE4-4C5E-4F14-806F-92DF54950A03@ceruleansystems.com> On Oct 25, 2009, at 1:18 PM, Stephen Williams wrote: > > It's all about psychological effects isn't it? These are often > indirect. The optimism of doing something via the stimulus somewhat > offset the cries of doom, gloom, and disaster long enough for people > to take a breather and get back to living and working more or less > normally. (Except the unemployed, who probably have already passed > through the angst of financial disaster and have found a fall-back > position.) The stimulus fell flat because it is predicated on everyone uncritically buying into transparently bogus magical thinking. Instead, they rightly dismissed the supposed upside as non-existent and accounted for the adverse consequences of epic wastefulness. Apparently there are not enough credulous Americans around for neo- Keynesianism to work as advertised. Neo-Keynesianism is the idea that you can astroturf your way to prosperity. Just because it is attractive to the political class does not mean it is a good idea. > The thing that I don't like about the stimulus choices is that they > gave away too much money with too little direct positive benefit. What did you expect? It is more than a little pollyanna-ish to expect that this money would not be largely wasted on political cronies. Even worse, a lot of R&D funding is currently being held up in Congress and will not be disbursed until some time in 2010. A lot of existing R&D programs are being suspended for lack of funds while billions are being liberally spent on pointless make-work programs that will never produce anything of value. > For instance, I would have favored spending a whole lot to bootstrap > a new auto industry, more or less as a clean break from the existing > one. Why? There are plenty of competitive auto companies already doing this. A waste of resources. Let's do something original instead. > Start serious efforts to create inexpensive "company town" like > environments, except not actually owned by the companies involved. > Perhaps something between a homeowners association and a credit- > union like structure. Why? These kinds of top-down initiatives have been tried numerous times before and failed just as often. We'd be better off with some original attempts at innovation rather than rehashing the same old, dead horses. > A key set of ideas would be to make this a push for highly capable > robotic factories, cheap, modular vehicles, electric or other low- > maintenance hyper efficient propulsion, and AI automated self- > driving (a la Darpa Challenge) vehicles. Resolve to solve the > technical, economic, and legal issues. After success, plan to solve > the efficient auto / train integration and automated aircars. Most of this is already under development and funded. Throwing even more silly money at it won't make the research happen faster, it will just waste money. > We have plenty of technology to get started on all of that, probably > having success in 5 years. We just have to collectively decide to > do it. We don't need to "collectively decide" a damn thing, except may be for everyone to get out of the way of people that *will* develop this technology. Collective decisions are good at producing exactly one thing: expensive white elephants that only a government could afford to waste money on. A fast path to future technological prosperity by political committee? Seriously? I thought we outgrew that silly idea in the mid-20th century. From sdw at lig.net Sun Oct 25 20:17:46 2009 From: sdw at lig.net (Stephen Williams) Date: Sun, 25 Oct 2009 20:17:46 -0700 Subject: [FoRK] Cutting through the stimulus BS In-Reply-To: <95CB8FE4-4C5E-4F14-806F-92DF54950A03@ceruleansystems.com> References: <8E4B57F8-639B-46BE-8EA2-B9AB98437D0C@place.org> <4AE4B294.90509@lig.net> <95CB8FE4-4C5E-4F14-806F-92DF54950A03@ceruleansystems.com> Message-ID: <4AE514DA.9010207@lig.net> You are assuming I mean one mechanism because you think I'm invoking an old pattern. J. Andrew Rogers wrote: > > On Oct 25, 2009, at 1:18 PM, Stephen Williams wrote: >> >> It's all about psychological effects isn't it? These are often >> indirect. The optimism of doing something via the stimulus somewhat >> offset the cries of doom, gloom, and disaster long enough for people >> to take a breather and get back to living and working more or less >> normally. (Except the unemployed, who probably have already passed >> through the angst of financial disaster and have found a fall-back >> position.) > > > The stimulus fell flat because it is predicated on everyone > uncritically buying into transparently bogus magical thinking. > Instead, they rightly dismissed the supposed upside as non-existent > and accounted for the adverse consequences of epic wastefulness. > Apparently there are not enough credulous Americans around for > neo-Keynesianism to work as advertised. > > Neo-Keynesianism is the idea that you can astroturf your way to > prosperity. Just because it is attractive to the political class does > not mean it is a good idea. Agreed. > >> The thing that I don't like about the stimulus choices is that they >> gave away too much money with too little direct positive benefit. > > > What did you expect? It is more than a little pollyanna-ish to expect > that this money would not be largely wasted on political cronies. > > Even worse, a lot of R&D funding is currently being held up in > Congress and will not be disbursed until some time in 2010. A lot of > existing R&D programs are being suspended for lack of funds while > billions are being liberally spent on pointless make-work programs > that will never produce anything of value. Painfully stupid. >> For instance, I would have favored spending a whole lot to bootstrap >> a new auto industry, more or less as a clean break from the existing >> one. > > > Why? There are plenty of competitive auto companies already doing > this. A waste of resources. Let's do something original instead. I don't see that there are plenty of well-funded competitive companies making much progress on these things. There is a little. > >> Start serious efforts to create inexpensive "company town" like >> environments, except not actually owned by the companies involved. >> Perhaps something between a homeowners association and a credit-union >> like structure. > > > Why? These kinds of top-down initiatives have been tried numerous > times before and failed just as often. We'd be better off with some > original attempts at innovation rather than rehashing the same old, > dead horses. I didn't mean that it was top down? Give 100 year leases on some government land or work out something for large tracts of cheap land, pledge to provide basic infrastructure (roads, utility long-haul), perhaps a little American Indian reservation-style financial guarantee arrangement, and allow entrepreneurs to start cranking. Company towns are more top-down than I am proposing. And they have poor dynamics, two reasons I was steering around them. > > >> A key set of ideas would be to make this a push for highly capable >> robotic factories, cheap, modular vehicles, electric or other >> low-maintenance hyper efficient propulsion, and AI automated >> self-driving (a la Darpa Challenge) vehicles. Resolve to solve the >> technical, economic, and legal issues. After success, plan to solve >> the efficient auto / train integration and automated aircars. > > > Most of this is already under development and funded. Throwing even > more silly money at it won't make the research happen faster, it will > just waste money. I still disagree about that. More researchers with more resources and incentivized people ready to build funded solutions that are created has to be better than a smaller R&D and entrepreneurial market. > >> We have plenty of technology to get started on all of that, probably >> having success in 5 years. We just have to collectively decide to do >> it. > > > We don't need to "collectively decide" a damn thing, except may be for > everyone to get out of the way of people that *will* develop this > technology. I'm advocating "get out of the way" plus "make way for" the people that "will" develop this technology. As an obvious example, the solutions to the Darpa Challenges would not have happened as soon or been as widely appreciated had there not been a Darpa Challenge. > Collective decisions are good at producing exactly one thing: > expensive white elephants that only a government could afford to waste > money on. > > A fast path to future technological prosperity by political committee? > Seriously? I thought we outgrew that silly idea in the mid-20th century. We have, mostly, government by committee. Government decisions are needed sometimes to cut through red tape, make something more attractive, or similar tweaks to the status quo. It can be useful, and sometimes not all that expensive. There are many better ideas than blindly bailing out the financial industry. sdw From andrew at ceruleansystems.com Sun Oct 25 21:55:05 2009 From: andrew at ceruleansystems.com (J. Andrew Rogers) Date: Sun, 25 Oct 2009 21:55:05 -0700 Subject: [FoRK] Cutting through the stimulus BS In-Reply-To: <4AE514DA.9010207@lig.net> References: <8E4B57F8-639B-46BE-8EA2-B9AB98437D0C@place.org> <4AE4B294.90509@lig.net> <95CB8FE4-4C5E-4F14-806F-92DF54950A03@ceruleansystems.com> <4AE514DA.9010207@lig.net> Message-ID: <4959ADC8-451A-484F-85CB-1EB889E0F45A@ceruleansystems.com> On Oct 25, 2009, at 8:17 PM, Stephen Williams wrote: >> Most of this is already under development and funded. Throwing even >> more silly money at it won't make the research happen faster, it >> will just waste money. > > I still disagree about that. More researchers with more resources > and incentivized people ready to build funded solutions that are > created has to be better than a smaller R&D and entrepreneurial > market. Multiplying the monkeys will not magically generate research results. And we most certainly cannot build a solution today, multiple required technologies simply do not exist and not for lack of research effort. We already spend a lot of money on this from multiple directions. It is more of a "insufficient really bloody smart people" than a "insufficient money" problem. Unfortunately, "really bloody smart people" are already fully employed. A lot of technologies are bottlenecked on multiple hard, theoretical problems that there are only a handful of people in the world qualified to even attempt to solve. And usually those people already are working on the problem. Developing a larger pool of people that have the talent and knowledge to legitimately contribute something useful will take years. You can't just pull people off the street, slap a "researcher" sticker on their forehead, and expect results. It takes years for even very smart people to build up the background in a subject area to make useful research contributions. We are already in the position where there are four nominally qualified deadweight research programs for every productive one. I do not see how adding a bunch of even weaker research teams will improve the productivity, though I can certainly see how it would waste a lot of money. In many cases, it would be a very poor investment. > I'm advocating "get out of the way" plus "make way for" the people > that "will" develop this technology. As an obvious example, the > solutions to the Darpa Challenges would not have happened as soon or > been as widely appreciated had there not been a Darpa Challenge. The DARPA challenges illustrate the basic defect of the model you are proposing -- which they were attempting to work around. First, that funding is completely subject to political whim and many R&D projects have been killed for purely political reasons without any regard for the value of the research. Second, what the DARPA Challenge demonstrated specifically is that much of the R&D funding has been captured by the politically connected without regard for effectiveness. DARPA has been spanked by Congress many times. I'm waiting for your solution to these fundamental problems with government-funded R&D. I do not object to funding useful R&D this way, but you simply want to expand and throw money at a deeply broken system. Also, a lot of important research problems are not amenable to the Grand Challenge model in any case, nor easily packaged into a concept a normal technical person could even grok. Many critical theoretical problems are the research domains of people who've spent many years on the problem, a population you can count on one hand. > There are many better ideas than blindly bailing out the financial > industry. Yet as you can see, that *is* the priority of the government. Forget hypothetical solutions, show me one that is actually implementable.