183Re: Grand challenge: Language evolution (Re: Autonomic email client?)
- Dec 1, 2001The problem with biology-inspired, or more general,
complexity-science-inspired approaches is that you don't know where they
If we undertake a project, we want a result like a self-correcting editor.
Biological systems, in turn, don't evolve to achieve a purpose, they evolve
into whatever survives and fits. As a consequence, you can't develop
systems in a traditional goal-driven or requirements-driven way, at least
unless you have found a way to make the construction-oriented approaches of
biological systems converge according to your will. But that is against its
grain and you would ha
I"m quite certain that a complexity-science solution to a self-healing
editor that corrects your writing will not have a traditional parse tree at
its heart. I can't tell you how it will look like, but if Dick had a hand
in getting it started, it might at times decide to add some irony to your
At 06:43 PM 11/30/2001 -0600, bhaugen wrote:
>May I suggest that the paradigm (forgive the word) is wrong.
>Don't think about software as autonomous agents
>(an artificial intelligence concept) think of software
>as intelligence amplification - extension of
>human intelligence - or intelligent assistants.
>Keep the humans in the loop.
>Make everything work better for humans.
>What would make our lives better?
>More humane? More beautiful?
>More lovely? More harmonious?
>How do we want to live?