Three Places Beberg Is Wrong

Jeff Bone
Wed, 02 Jan 2002 00:23:55 -0600

Okay, here're three very general (capability, not application) areas
off the top of my head that are going to be hugely lucrative in IT
over the next five, that will continue to drive demand for cycles,
bits stored, etc. --- and will directly or indirectly transform the
way we work, play, and interact.  This isn't at all an exhaustive
list, just three areas I happen to have been looking at in some
detail for a while now.

INTEGRATION.  This is huge.  Even though we've got a critical mass of
systems "on the Web" in some form or fashion, we're still largely
"balkanized" in some sense.  That's problematic when you have
business processes, decisions, etc. that need to coordinate between
and synthesize data / subprocesses within multiple organizations.
We've got a bazillion connectoids, directories, metadata structures,
interfaces and interface technologies, etc.  Wiring all this stuff up
is going to take somebody's time and somebody else's money.  My bet,
of course, is on REST + event infrastructures.  This past summer plus
a continued lurk on soap-builders and elsewhere has convinced me that
the Baker Argument is fundamentally correct.  The beneficiaries here
will be a *very* few infrastructure vendors and *lots* of PSOs, and
the work will be in two waves:  the first wave of attempted "Web
Services" implementations followed by later, more loosely coupled
implementations a few years out when the scalability issues are fully
recognized.  (This tracks to the evolution and deployment of
client-server stuff 10 or so years ago, if you recall.)

STORAGE.  The economics of storage are fundamentally broken for all
kinds of reasons.  We're already at / beyond a crisis stage in
certain industry segments, and the accumulation of stored information
over time is a key driver;  don't bother telling me how drive prices
are going to zero, as that's not the solution --- it's part of the
problem.  Nor does consolidation (SAN/NAS) address the key problem
set.  Capacity increases, but utilization of capacity remains flat,
which means information grows, and management costs are proportional
to data stored.  This is one I feel pretty close to these days. ;-)

INTELLIGENCE.  Information isn't power, intelligence is.  We've
created a huge and growing problem simply by lowering the cost of
information creation and dissemination so much.  Huge amounts of data
are available to anyone with just a few keystrokes, almost for free.
We've minimized the artificial frictions of data access, now we're
flooded.  Extracting useful information out of this flood --- the
right information at the right time --- is a huge problem, one for
which solutions will be lucrative in numerous business domains.  And
IMO "the Semantic Web" isn't the answer;  it's a hammer, but not
everything is nails.  The ability to process and act on
"unstructured" information in interesting ways is going to be huge.

Just some thoughts,