He talked about Harold Cohen’s AARON http://en.wikipedia.org/wiki/AARON, a very impressive program that creates artistic images, and has been under development and improving for 35 years. AARON was originally written in C, but was rewritten in Lisp in the early 1990’s.

He talked about the first (or, an early) Lisp compiler, written by Michael Levin and Tim Hart. He told the story of Josetph Weitzenbaum’s http://en.wikipedia.org/wiki/Joseph_Weizenbaum famous ELIZA program http://en.wikipedia.org/wiki/ELIZA. Kunze wrote a version of Eliza that he put on the web (it’s gone now) with 10,000 questions, which got a great deal of use: over 682 million interactions.

He then talked about Lisp’s current unpopularity, and claimed the reason for it is that Lisp programmers all have Asperger’s Syndrome. His supposed evidence for this is a short online quiz that some programmers took.

The quiz came from Simon Baron-Cohen http://en.wikipedia.org/wiki/Simon_Baron-Cohen, and much of what Kunze said seems to come from Baron-Cohen’s, such as his theory about autism-spectrum disorders being “an extreme form of the male brain”. Except that Baron-Cohen refers to these things as “hypotheses”, whereas Kunze cites them as facts.

– The idea that you can dianose Asperger’s with such a quiz is preposterous.

– If he is referring to Simon Baron-Cohen’s EQ SQ tests http://eqsq.com/eq-sq-tests, they do not claim to diagnose Asperger’s syndrome.

– I have had extensive experience with people who have Asperger’s, and am absolutely certain that the vast majority of Lisp programmers do not have it.

– I asked whether he had tried the quiz on non-Lisp programmers, and he admitted that he had done so with the same results, even though he had been very specifically directing his attacks to the Lisp community specifically.

He said that you can tell that they have Asperger’s Syndrome because they hate to travel, and don’t like being with many other people. I felt this was ironic, since he was talking to a room full of Lisp programmers who had travelled in order to be with a lot of other people.

He also accused Lisp programmers of being unable to compromise, and being unmanagable. This is the opposite of my own experience, and I wonder if it says more about his own experiences as a manager than it does about Lisp programmers.

He said that languages take 15 years to mature, citing 1984-2008 for Common Lisp, and 1995-2008 (not yet mature) for Java. He didn’t say how he came up with this.

He said that a big problem was that there were not enough trained people, citing a figure of 900 computer science Ph.D.’s in the 1990’s, and 1000 now. Why he thinks that being “trained” requires a Ph.D., he did not explain. He also said that he feels it takes four years to become a good programmer, but that you “cannot” retrain people: for example, at Electronic Arts, 60% of the developers could not move from 8-bit to 16-bit processors. This seems quite implausible to me; he didn’t say where he got this figure.

We can just write applications in Lisp to compete with other applications, being orders of magnitude “better”. Applications with less bugs, applications more intelligent, etc.

If you want to improve Lisp’s place in the market, just write more Lisp applications.

I actually tried taking a free Aspergers quiz online (I can’t remember where now), since I had heard speculation about how they correlate, and got a result of:

Aspie score: 73 out of 200
Neurotypical score: 129 out of 200
with “very likely neurotypical”

In the breakdown of different mental aspects though I scored “above average” on the “Aspie” score and “below average” on my NT (neurotypical) score on a few of them.

If the company culture as whole absorbs too many “aspergerisms”, then the business ceases to be able to talk to its customers!

LOL! Well that explains right there why companies don’t do this, then.

Maybe I just don’t understand Lisp, but I don’t agree with JohnL. As far as “changing Lisp’s place in the market” I agree there’s nothing that can be done in the short term, but I go back to what Alan Kay and even what Peter Seibel has said in the past: it’s about culture. Folded into that is it’s about a way of thinking, but I don’t think you have to “give people Aspergers” for them to get it. For example Richard Stallman talked about how one of his associates wrote the first Lisp implementation of Emacs on MULTICS, wrote an easy to follow manual for it, and gave it to his organization to use, and the secretaries were modifying Emacs in Lisp. The key he said was “don’t call it programming”, because that scared people away from it. They learned programming without realizing it, or put another way, “while doing something else.”

Personally I found JohnL’s response kind of funny, from your description. As McCarthy said, just as before, people can go off and create their own Lisp if they want. Nothing’s stopping them. I guess what JohnL was really saying was, “If you’re going to go off and do this, don’t expect any support from the likes of me.”

Kay has complained, though not bitterly I think, that Smalltalk and Lisp “eat their own young”. What he’s meant by this is they are both so useful that most people who get into them don’t aspire to create anything better, and this contributes to the slow progress of CS research. He’s said that using building architecture as an analogy, Smalltalk is at the level Gothic architecture (1,000 years old). Current software development practice with current tools is at the level of ancient Egyptian pyramid building (4,500 years old). I haven’t heard him give a historical analogy to Lisp. He has said something like “It’s the most beautiful language ever written.” This gives you an idea, though, of how far he thinks we have to go before CS gets to the level of modern engineering.

Putting defmacro in your language is a way of saying: “Wheee! language design is fun. Here, you have a go!” It is about giving up control.

Ofcourse, some people cannot abide parentheses, so Lisp has read macros, to yield to their wishes also.

“We were there when Lisp was actually useful,” he declared. “Nothing we can do today can change Lisp’s position in the market.” My notes are sketchy, of course, but that’s the gist of it. I was stunned….