4 ms·

I'm really surprised out of current nearly 300 comments and counting, nobody really is mentioning constraint programming or CP, a seemingly forgotten determinis

I'm really surprised out of current nearly 300 comments and counting, nobody really is mentioning constraint programming or CP, a seemingly forgotten deterministic sibling of the stochastic data-driven AI [1],[2],[3]. "Out of these 14 tasks, we guessed that only 6 require abstract reasoning alone to perform. Strikingly, we classified only one of the top five most important tasks for medical scientists as relying solely on abstract reasoning. Overall, the most critical aspects of the job appear to require hands-on technical skills, sophisticated coordination with others, specialized equipment use, long-context abilities, and complex multimodal understanding." Almost all of 14 R&D tasks listed in the table including that are not suited for data-driven AI with abstract reasoning can be solved by CP. Provided that we are allocating enough compute resources with at least similar to that we're currently providing data-driven AI with the crazy amount cloud networks of massively parallel compute CPU/GPU/TPU/etc. Fun facts, the modern founder of Logic, Optimization, and Constraint Programming is George Boole, the grandfather of Geoffrey Everest Hinton, the "Godfather of AI" and "Godfather of Deep Learning". [1] Constraint programming: https://en.wikipedia.org/wiki/Constraint_programming https://en.wikipedia.org/wiki/Constraint_programming [2] Logic, Optimization, and Constraint Programming: A Fruitful Collaboration - John Hooker - CMU (2023) [video]: https://www.youtube.com/live/TknN8fCQvRk https://www.youtube.com/live/TknN8fCQvRk [3] "We Really Don't Know How to Compute!" - Gerald Sussman - MIT (2011) [video]: https://youtube.com/watch?v=HB5TrK7A4pI https://youtube.com/watch?v=HB5TrK7A4pI

hyperturtle 2y ago
I've always wondered if it would work out well for an LLM to evaluate a constraint program/script when it encountered a problem involving constraints / logic. like or-tools with python or evaluating a minizinc program.
- teleforce 2y ago
  Yes, that's very plausible and I've thought similarly but not exactly the same approach for the diagramming limitations of LLM [1]. Fun facts, the presenter John Hooker was asked about how to determine the suitability of a particular heuristic solver for specific problems in his presentation. He casually answered that if he knows the solutions to that he will probably win a Nobel Prize. But perhaps AI/LLM can help in a way to recommend the solver based on the type of applications or problems. If I'm not mistaken there's also Donald Knuth (TAOCP) asking questions after the JH's presentation, how often you see that? [1] Diagrams AI can, and cannot, generate: https://news.ycombinator.com/item?id=43424916 https://news.ycombinator.com/item?id=43424916
babyent 2y ago
I’m using constraint programming in my platform. Any experts want to help me?

hyperturtle 2y ago

I've always wondered if it would work out well for an LLM to evaluate a constraint program/script when it encountered a problem involving constraints / logic. like or-tools with python or evaluating a minizinc program.

teleforce 2y ago
Yes, that's very plausible and I've thought similarly but not exactly the same approach for the diagramming limitations of LLM [1]. Fun facts, the presenter John Hooker was asked about how to determine the suitability of a particular heuristic solver for specific problems in his presentation. He casually answered that if he knows the solutions to that he will probably win a Nobel Prize. But perhaps AI/LLM can help in a way to recommend the solver based on the type of applications or problems. If I'm not mistaken there's also Donald Knuth (TAOCP) asking questions after the JH's presentation, how often you see that? [1] Diagrams AI can, and cannot, generate: https://news.ycombinator.com/item?id=43424916 https://news.ycombinator.com/item?id=43424916

babyent 2y ago

I’m using constraint programming in my platform. Any experts want to help me?