alt Hacker News> "Even though I can motivate it in retrospect, ChatGPT’s idea to use h^2-dissociated sets to control relations of order at most h feels quite ingenious. As far as I can tell, this idea is completely original."
The question that keep bothering me is can an LLM generate an idea that is truly novel? How would/could that actually happen? But then that leads to the question - what are we actually doing when we think?
Perhaps it's as simple as the ability to just make mistakes that matters, the same things that powers evolution. As long as the LLM can make mistakes, it's capable of generating something genuinely novel. And it can make more mistakes much faster than we can.
Replies
Trivially the answer is yes by the infinite monkey theorem. If we allow the sampler to pick any token then any stream of arbitrary tokens can be generated. Therefore if an original idea can be represented with written words then a LLM can generate it. That is perhaps not the most satisfying answer, but if you want a better one you'll need to provide a function that determines if an idea is original.
For my paper about ME/CFS, I let an LLM integrate lots of findings of other scientific papers. Then I ask the LLM to "creatively brainstorm", given all we know of ME/CFS and the newly integrated paper, to generate new hypotheses, treatment ideas or any other kind of insight it can think of.
This works really well.
Now, it's clear that I have no idea how much of this is something we would consider new and original, and how much is a kind of systematic, but not novel, easy of thinking.
What I couldn't do so far is get an LLM to generate a truly new maths theory, with new abstract concepts and dimensions and points of view. The kind that is not just a combination of existing theories and logic.
My own take, and it's veering into the Philosophy of Mathematics, but there's a debate about whether Mathematics is "Invented" or "Discovered".
If it's "invented", then it requires ingenuity.
If it's "discovered", then it was always already there, just waiting for the right connections to be made for it to be uncovered and represented in a way we can understand.
Invention requires ingenuity, but discovery does not. So if LLMs can generate truly novel mathematics, for me that settles it that mathematics is indeed discovered, as LLMs are quite capable of discovery yet I don't consider them possible of invention.
It's about the ability to combine ideas in novel ways, without breaking the rules in relevant frameworks. Sometimes the idea may even be to contradict existing theories where they are weak.
How do you define a new idea?
To me, it's rearranging the information you had in a way that hasn't been applied or published before.
That's literally what LLMs are built for.
Yes, they can.
Some people like to parrot "next token prediction", "LLMs can only interpolate", and other nonsense, but it is obviously not true for many reasons, in particular since we introduced RL.
Humans do not have the monopoly on generating novel ideas, modern AI models using post training, RL etc can come to them in the same way we do, exploration.
See also verifier's law [0]: "The ease of training AI to solve a task is proportional to how verifiable the task is. All tasks that are possible to solve and easy to verify will be solved by AI."
This applied to chess, go, strategy games, and we can now see it applying to mathematics, algorithmic problems, etc.
It is incredibly humbling to see AI outperform humans at creative cognitive tasks, and realise that the bitter lesson [1] applies so generally, but here we are.
[0] https://www.jasonwei.net/blog/asymmetry-of-verification-and-...
[1] http://www.incompleteideas.net/IncIdeas/BitterLesson.html