- cross-posted to:
- [email protected]
- [email protected]
- cross-posted to:
- [email protected]
- [email protected]
You must log in or register to comment.
So remembering is not thinking?
Does it matter if we get more math problems solved?
Yes, it obviously matters if we have more math problem solved because a lot of math problems directly apply to engineering which lets us build new and interesting things we couldn’t make before.
Apologies, meant to be taken in the context of the headline:
Does it matter that AI is remembering if on the whole we are getting problems solved faster?
Oh I agree there, the end result is what matters. And I think we can view this in a similar way to a person using a calculator to crunch large numbers. It’s just a tool to help us think about bigger problems.
This is a problem for me. I’m a good rememberer but not a great creative thinker.
I think this is an interesting thread. Is creativity just a matter of combinatorics over the right data? Or is it something more? Like you are thinking about existing solutions and theorems, and then you trip over a bucket and come up with bucket sort or something
I think this distinction is imagined, thinking is largely just a reorganisation and reinterpretation of memory.
And that reorganization and reinterpretation is the difficult part for AI.
But my point is thinking is all smoke and mirrors, its really just an active memory.
I disagree, thinking is simulating. Our brains evolved to create an internal world model which allows us to simulate hypothetical events and project into the future based on that. Abstract thinking is just applying this mechanism to non-physical things such as math, where we internalize the rules of the system through study and then develop intuition for creating valid connections.
I largely agree with what you said, the problem is phrases like ‘simulate hypothetical events’ and ‘applying this mechanism’ are just other ways of saying thinking and do not pin down what thinking is. I know it is not understood what thinking is so I cannot expect you to solve it on Lemmy. I do think it is the manipulation and re-weighting of memory and did therefore feel the original statement that AI models have good memory but poorer ability to think was a little tautologous.
The success of LLM in mimicking human communication makes me wonder if they have unintentionally solved what thinking is and that the human mind is probabilistic. Our minds might decide upon the next event by experience of what successfully followed in similar previously observed situations. This probability work all being run in the subconscious mind, the answer then being passed to the conscious mind which congratulates itself on thinking it out. Similar systems work for picking up a cup, my conscious mind decides I want to pick up a cup but it is the subconscious mind that does all the hard labour of sending appropriate signals to the muscles in my arm. As with reasoning this could have been learned by the subconscious probabilistically. In infancy sending out many signals and seeing which reached for the cup.
For sure, we don’t really have a formal definition for what things like thinking or consciousness are right now. So, it’s all a bit slippery as a result. :)
I do think that our minds are probabilistic engines, and what makes the whole thing work is the feedback loop. Since we have embodied consciousness, our internal simulation of the world gets constant feedback from primary reality keeping it on track. And we can see how quickly human minds go off the rails when that loop is broken when people get disorders like schizophrenia.
I kind of view consciousness as a workspace for the mind with most of the work happening within the subconscious. Consciousness is a mechanism which forces all the parallel processes within the mind to go through a single bottleneck and settle on a single course of action.
Incidentally, I really enjoyed The Society of Mind by Minsky, I am A Strange Loop by Hofstadter, and Consciousness Explained from Dennett. They do a fairly good job discussing the plausible mechanics.
This distinction gets explored some in the book Children of Ruin.
I enjoyed it quite a bit. :)
That’s a neat hypothesis, but there is no settled science on what is “thinking”. Specially thinking in the sense of coming up with new ideas.
True, but if it is not known what thinking is then it is hard to be sure AI models are not doing it.
Well yeah… AI research started out in the Soviet Union precisely as an investigation of intelligence and consciousness. The goal was never to replace workers (as you know, the USSR was socialist). It was a scientific investigation of our own minds. The idea was that by learning how to build an artificial mind, every step would help us learn about our own “natural” minds.
It’s not random that some of the best existential works of sci-fi ever came from the USSR 😂
