>>106175357
>Reasoning is literally an emergent behavior of LLMs.
Can you show me any evidence whatsoever of LLMs "reasoning"? Even explicit attempts to fake reasoning using CoT hacks ends up failing as in pic related. Your emergent reasoning fantasy is quite literally schizophrenia.

>>512323108
>>Source
>>An AI company trying to sell its AI
Yep. They literally set their own fake goalposts and then train bots to surpass those goalposts. They've been doing this shtick every month, for years. "AI surpasses humans at X [according to our benchmark]" in an endless loop, then you actually try to use it and end up with pic related results.

>>512226064
If your chatbot is so smart, why does it get mind-broken by simple logic puzzles?

>>512047305
I accept your concession that there's 0 evidence anyone has ever created a successful product using """AI""" slop. Your anecdotes on the internet count for exactly nothing.

>>16728975
>no recursive self-improvement
>singularity
Another episode of tards not understanding the words they use.

>it has more accumulated knowledge
It hallucinates constantly.

>can think faster than 99% of the world's population
It can't think at all.

>>510710610
I quit professional programming long before """AI code""" became a thing. I have no personal fucks to give about wagies potentially getting replaced. It just rubs me the wrong way how marketing kikes keep trying to gaslight everyone into the idea that """AI""" can do anything that requires thought.

>>510561380
>llms will do it better
In exactly two more weeks. Meanwhile in reality pic related happens whenever they have to deal with something slightly outside the training set.