Mia Golec, Emily Davis (both of CMU), and I have been spelunking in Simon and Newell's archives at CMU and came upon a remarkable complete listing of Simon's "heuristic coder" (https://dl.acm.org/doi/pdf/10.1145/321186.321192) [in the paper he calls it a "heuristic compiler", but the listing uses the term "coder"]. I've only just begun studying this fascinating find, which I intend to try to run on one of our IPL interpreters!
I know I'm not supposed to follow up my own posts, but this is truly an amazing update! We (being Claude and I) just got Simon and his Daughter's 1962 IPL-V version of Victor Yngve's famous (1961) random sentence generator (originally in COMIT), to run on Paul Kimpel's amazing IBM 1620 emulator using Beyer and MacDonald's 1963 IPL-V comterpreter (from U Oregon)! If you don't want to wade through all the running of it yourself (which, BTW, I highly recommend, it's AWESOME TO SEE IT ACTUALLY RUN!) you can see the output at: https://github.com/jeffshrager/ExecutableArchaeology/blob/ma...
I was an (adjunct) prof at Stanford for over 20 years. I taught the top level interaction analysis course - basically cognitive science applied to human interaction with complex systems, including AI based system. As soon as I saw students turning in AI papers and projects, I knew all of upper education as we know it was fucked. I retired. Education is fucked, but learning isn’t! This will be a golden age where you can learn anything you want to. If you don’t want to learn anything, you’ll be a meatbot working for those who do. Many people may be happy in that role. I know someone who is extremely happy working at as a barista. He was also happy having gone to college for a useless degree. He’s happy he did that. He tells me the things he’s learning on his own every week from AI tutors. Quantum Mechanics. People will find a way to make their way, mostly. Education as currently organized is fucked. Teachers should be there to help students learn what they want to learn (at least after elementary school, which is still important and won’t be disintermediated). Like flight instructors. People who want to learn to fly learn most of it on their own, and the flight instructor is there to help (and keep you safe at least at first, until you solo). People seek out flight instructors, and will continue to do so (at least until airplanes fly themselves - but that’s a different discussion). Teachers will become tutors. And I dare say that teachers will like it better than their current role of having to force-feed 2/3rds of a class who doesn’t want to learn. Instead they’ll enjoy helping the few students who actually care to fly to fly.
AI is great for those who want to learn. Earlier, one has to find all relevant books, search specific forums, or use google in order to find answers to questions that teachers or fellow smart-students find silly. Even if AI gives false answers, it is better than having no answer. At least, the search space to find an answer is reduced.
Yes. All of those were written in IPL well before Lisp. There were a dozen IPL-V implementations across almost every common computer of tat time. (Of course, there weren’t that many common computers! :-)
True enough that IPL inspired Lisp (along with Church’s Lambda Calculus), but, to my mind, what is interesting about IPL-* is that the first true AIs and Cognitive Models were written in it/them, esp The Logic Theorist, which was the first heuristic search program, and which actually improved upon proofs by some of the most famous logicians of the time. David Moews has made several of the original IPL Logic Theorist programs run again (although not on this 1620 platfor’. . . Yet!) :: https://news.ycombinator.com/item?id=48116935
Paul Kimpel has built a lovely web-based IBM 1620 emulator, and was able to compile and run a 1963 IPL-V compiler and runtime from U Oregon. Run it all from your browser! Execution details and project history linked from Paul’s repo: https://github.com/pkimpel/retro-1620
Earlier this week, having come across my ArXiv paper about reanimating the Logic Theorist, David Moews (https://djm.cc/dmoews.html) wrote me about an amazing piece of work he's done:
He built an interpreter for the original IPL-I pseudocode of the original Newell and Simon Logic Theorist, straight out of the 1956 RAND report (P-868), and then got it running!
(I'll call the version that David reanimated "LT1" or "LT56", and mine "LT5" or "LT63" because mine was rewritten for IPL-V and published in 1963.)
What makes David's work especially interesting, aside from pushing the RetroAI window back 8 more years (!), is that IPL-I was NEVER ACTUALLY IMPLEMENTED! It was hand-executed by Simon's students (and supposedly his kids!) simulating the imagined IPL-I machine. This actually makes the problem much simpler. (Not at all to diminish David's accomplishment!)
In the 1956 report Newell and Simon describe the process in something close to the cognitive operators they hypothesized underlay human theorem proving. This is essentially LT1: A (somewhat) high-level specification of the Newell and Simon theory of cognitive theorem proving. But because LT1 didn't have to actually run on a real computer, it could depend upon human intelligence and flexibility to handle the complexities of actual implementation that you need to do to make a real computer actually do the whole thing end-to-end. (Or, as in David's case, a pile of Python code, which, of course, Simon and Newell didn't have in the mid 1950s!) As a result, LT1 is a bit over 400 lines whereas LT5, which is what you get when Shaw had to actually nail down the complexities of actual implementation, is nearly 3000 lines!
Anyway, huge congratulations to David; well worth a look if you care about the prehistory of AI, Lisp, or theorem proving. His repo is here: https://github.com/dmoews/logic-theorist. The readme provides a lot of intersting and important detail that I've glossed over.
Lars Brinkhoff brought up PARRY, Kenneth Colby's 1972 paranoid chatbot, and Rupert Lane, who recently reanimated the original ELIZA, was able to (sort of) reproduce the famous RFC439 trans-internet conversation between ELIZA and PARRY. See the blog post for the amazing (and amusing) details! (Turns out that RFC439 used a Lisp ELIZA, not the MAD-SLIP original.)
"Inspired by true events, this world premiere by award-winning playwright Tom Holloway is a gripping psychological thriller about the birth of AI, and the ethical consequences of outsourcing our humanity."
reply