Re: software survival and progress (was Re: Sweeden & Germany to phase out nuclear power?)

From: Mike Linksvayer (ml@gondwanaland.com)
Date: Sat Mar 30 2002 - 18:27:11 MST


On Sat, 2002-03-30 at 12:10, Samantha Atkins wrote:
> It is just a guess as is mine. I certainly know of a lot of
> very good brains currently not being gainfully employed and a
> lot of others that are in a holding pattern doing everything
> necessary to keep a company afloat rather than doing what they
> do best. The company I worked for was cut down to and arguably
> into the bone. We are still alive but not as able to seize
> opportunities and flesh out our technology as before.
 
OTOH I know a number of good brains that have gone back to school, taken
up working on their own projects and/or traveling. These sorts of
things cause financial stress, but in the long run may better serve
creativity and innovation than a longer run at theoretically innovative
startups rapidly morphing into overweight bureaucracies dominated by
politics.
 
I wrote:
> > I guess one can quibble over what constitutes a significant advance or
> > claim that nothing's new under the sun, but what about (in no particular
> > order)

[...]
> > * easy internationalization and accessibility
>
> Not relevant to the problem.
 
It helps indirectly, by sparing cycles otherwise spent on these problems
and by increasing the number of people who can effectively use
computers, some of whom should be brilliant programmers. The
accessibility bit may even be useful for humans with augmented/altered
senses and motor devices.
  
> > I consider adoption and robust implementation highly relevant, even if
> > the ideas are decades old. None of the above have run their course
> > IMO. I look forward to another decade of improvements (though with a
> > half-empty perspective, one may be looking at a decade of stagnation)
>
> Not one of the things listed address better automation of the
> software process itself and radically better toools and
> techniques in programmers hands.

All of your critiques were insightful, but I think overly discount the
bright side: sure nothing I mentioned is new, but widespread
understanding and adoption is, and in my (again, rather limited)
experience each accrues productivity gains and frees one to concentrate
on higher level solutions that are just as hard as the lower level ones
we don't have to recreate. I would hate to go back to even 1995.
 
> Intentional programming is extremely interesting to me. In the
> meantime a great deal could be done to reverse-engineer and
> refactor existing code (providing much more useable information
> about the codebase as a side-effect!) and to enable much higher
> levels of reuse and use of design patterns. But creating a
> viable tools company was very difficult before the meltdown. It
> is even more so now. So it is likely these developments, if
> they happen, will come through volunteer efforts as Open Source.
 
I certainly wouldn't want to start a tools company now, or perhaps in
the last decade. Most of the world is wedded to Microsoft's tools
(there's a tools company, much as I dislike them) and much of the rest
demands libre tools. Perhaps one can have it both ways: create
radically better open source tools, and form the next Cygnus Solutions
around them. Collab.net's <http://tigris.org> is the closest I can
think of to this right now. The Free Software Business mailing list is
a good place to pick at these sort of ideas.

> The stack is getting much broader but it is not getting
> significantly "deeper" or better organized for excellent use.
> The very structure of many software companies and of software
> patents and other encumbrances precludes much reuse of prior
> efforts. A lot is re-invented over and over again by working
> software engineers. Relatively little is reused even within
> companies much less across them.

Certainly just as much code is rewritten as ever, even within companies,
but if you consider the broad and occasionally deep libraries that are
now available for any popular environment you've got massive code reuse
happening.

I understand that to you, a world class programmer with 20 years
experience who has worked with or independently discovered everything I
listed long ago, the software world may well appear stagnant. To me,
lots of cool things are happening. Software sucks in general and tools
in particular, though IMO a good deal of it "sucks less". Is the pace
enough to get us to the point where a seed AI can be built in time to
bootstrap a singularity in the current consensus (2020?) timeframe? I
don't have any insight on that question, and perhaps that's one reason
I'm not counting on a singularity.

What tools/methodologies do people think will be necessary to facilitate
seed AI engineering? If you had funding to hire 10 brilliant people for
the next several years, what tools would you create?

I'd focus on security, but that's just because I'm pain averse, and I
find the thought of an AI/upload world with today's security stack
incredibly scary. But I don't know that security that sucks less gets
us closer to a singularity. Perhaps closer to a desirable signularity
(mispeling unintended, but I'm glad I noticed it).

Mike Linksvayer
http://gondwanaland.com/ml/



This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:13:09 MST