[Edu-sig] As We May Think: What will we automate?

[Scott David Daniels]

kirby urner wrote:
>> At the other end, Python gives me a language I can talk to another
>> programmer in, and I can also run parts of the discussion on a machine.
>> There are other languages that do that, of course, but none that are
>> so easily communicated to a "random other" without spending more time
>> talking about the mechanics than about the idea.  I suspect this is why
>> Kirby likes APL so much, he can easily express large-swath ideas.  For
>> me, APL too quickly becomes terse little chunks.  But Kirby and I
>> program about different things.
>>
>> --Scott David Daniels
>> Scott.Daniels at Acm.Org
> 
> Yeah, plus when I got involved with APL in 1976-1977, we didn't have
> Python.  This was the first / only language with REPL in my reality,
> i.e. I could type at a terminal and get an immediate reply, what a
> difference!  Same think people like about Python.
> 
> My APL is rusty by now, so if someone wants to collaborate with me on
> communicating some large-swath ideas in at least partly working code,
> I prefer Python.  Like here's some "manga code" from the PPUG list:
> 
> http://mail.python.org/pipermail/portland/2009-March/000637.html
> 
> Thanks for you input Scott.
> 
> Kirby
You should definitely take a look at

http://arstechnica.com/science/news/2009/03/building-a-better-way-of-understanding-science.ars

as this is where Python can be very useful in science -- the
"stand at a whiteboard and scrawl and argue" phases.  I do it
for computer science, but I've used it to talk evolution with
a creationist -- explaining how recognizers can (and are) trained
to match things from weighted inputs and evaluate-crossover cycles
where the programmer has no idea how to solve the problem, but can
train a machine to do so.


--Scott David Daniels
Scott.Daniels at Acm.Org