0
$\begingroup$

Some years ago, I came across a notion that states something like: "a machine cannot leave its current level of thinking, move one level up and think about the level below", meaning that a machine cannot judge its lower-level deeds until we hard-code this ability for it. This meta layer thinking is what separates humans and AI.

I also remember I heard that as a rule of thumb, if we hard code 5 layers of meta-layer thinking for the machine, then its behavior would be satisfyingly humane, maybe enough to pass the Turing test.

I don't remember if it was a book, a documentary or a paper, and I cannot find any resource by searching key-words.

Does anybody know what exactly this theory is? And how can I find a good resource to read about that?

I also vaguely remember that it was related to this proven theory that "a machine cannot program another machine on its level, because it requires meta-layer thinking" or something like that.

$\endgroup$
9
  • $\begingroup$ It's difficult to say something if you don't remember anything about the source. We could come up with several answers, but that's not very productive. I encourage you to do a little bit of research and provide more context. For example, was it a "philosophical theory" or something like that? Was it in the context of AGI? $\endgroup$ – nbro Apr 18 '20 at 1:20
  • $\begingroup$ @nbro I did some searches. All I can remember beside the vague explanation, is the 'meta-layer' or 'meta-level' keyword. This keyword leads me to some motivational speech about thinking out of the box! But I somehow remember it was related to that proof that a machine cannot code another machine on the same level. Thanks for the editing btw! $\endgroup$ – Alireza Apr 18 '20 at 11:33
  • $\begingroup$ No. It's completely possible for an AI to code even better version of an AI. Layer of thinking is a very abstract and philosophical thing. It's useless. $\endgroup$ – RewCie Apr 27 '20 at 4:42
  • $\begingroup$ @abhas_RewCie I recall this "AI cannot do it" thing was a mathematically proven proposition based on Godel's Incompleteness theorem. I just cannot find related materials and papers. Would you please provide related literature that supports your comment?... and I don't get it when you say 'philosophical' and 'useless'. Useless how? are you suggesting that 'Philosophy of mind' is useless in the interdisciplinary field that is defined as the intersection of 'philosophy of mind/cognitive sciences/computer sciences'? $\endgroup$ – Alireza Apr 27 '20 at 9:39
  • $\begingroup$ @abhas_RewCie and just to give you a brief clue that many profound computer science frameworks have been ignited by the philosophical aspect of the researchers' studies, you can take a look at the concept of 'ladder of causation' in the Causal Inference field, or the definition of 'Counterfactuals' mentioned by Pearl, and take a look at his book 'the book of why' and how he has formulated the framework based on how human think about causation and exhibits 'creativity' and 'problem solving'. These abstract ideas are maybe useless for practical ML, but not for AGI $\endgroup$ – Alireza Apr 27 '20 at 9:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.