The hype cycle for Google’s fabulous new AI Co-Scientist tool, based on the Gemini LLM, includes a BBC headline about how José Penadés’ team at Imperial College asked the tool about a problem…
No, the typewriters are supposed to be random, this is guided based on previous work, so a whole space of output becomes extremely unlikely (so without looking at math for it, those spaces would show up very rarely if you then ran the infinite monkey experiment infinite times).
aren’t so-called AIs really just a computer version of the infinite monkey/ typewriter thought experiment?
No, the typewriters are supposed to be random, this is guided based on previous work, so a whole space of output becomes extremely unlikely (so without looking at math for it, those spaces would show up very rarely if you then ran the infinite monkey experiment infinite times).
It’s like that but if they were guided towards answers that sound reasonably likely