How is it possible that AI can't count r's in strawberry, but can solve the stonks problem?
>>109800330
>>109800330kys tranny
>>109800330My local llm has no problems counting R's
I am wondering what part of intelligence is tool use and which part is "integral" to intelligence itself.E.g. I can not calculate 34984583459 * 38459345083 in my head. Neither can LLMs.But an LLM can intelligencly construct a python script to calculate that and return me the result and I can open the calculator app on my phone and do the same.Now the AI can also call a minimal python script to count the number of r's in strawberry and return it to me with perfect accuracy.I, however, do not need to do that. I intrinsically know how to count them by simply looking at the word.Fascinating stuff.
>>109800345you first granny tourist.
Because it’s not counting
Counting r's is an extremely complex unsolved problem.
AI is just a markov chain generator. It does no thinking.
>>109800330>solve the stonks problemqrd?
The average man can’t tell the difference between pink and mauve yet men sent us to the moon and conquered the natural world.
>>109800330it's funny that back when strawberry was a secret codename everyone was trying to figure what super sophisticated search algorithm it could be referring to and it ended up just being chain of thought which is dead simple in comparison lol
>>109802538>men sent us to the moonRight
>>109800330Stonks more like stinks, am i right?
>>109800330they probably constructed an ai model specifically for figuring out the sonks problem. the r strawberry problem is general purpose ai i think. the larger models are probably dumber.
