/lit/ - Literature

>>23454885
I don't think AGI is that far away imo.
ChatGPT, image generation, applications of machine learning in medicine, engineering etc. are all techniques. They are like how a human has a Broca's area for speech and a primary motor cortex for initiating somatic movement. What separates us is that we have multiple techniques linked by a metacognition that chooses the correct one for the task (e.g. much as you don't ask chatgpt basic math calculations because it is inaccurate, the human brain doesn't use Broca's/Wernicke's areas to do math).
It is very feasible, at least to me, where you create a metacognition algorithm that calculates the minimum error based on technique selection for the task. At this point with good training data selection and enough computational power, you would essentially have AGI.

give me something about AI cults
thanks

>>23090428
>gets married and haves children
Maybe he's betting on human extinction being a transhuman transformation event instead of a genocide. Today's children could be tomorrow's Von Neumann probes shooting into the void in all directions to harness potential energy for the Bataillean death drive.

>>23040769
>>23040796
So when do we get to go full transhuman? I for one would like mechadendrites for... reasons.
>>23040804
You forgot the best part. One of the attendees found it so offensive they shouted, "Some of us are still Marxists, you know!", and left the room.
>>23040809
Why is he backsliding into regular racism after having transcended to the level of hyper-racism?
https://affirmativeright.blogspot.com/2014/10/hyper-racism.html

>>22997329
>It's like running multiple operating systems at one time.
The human brain is nothing like a Von Neumann machine. Even if it were, I would just make a smaller partition to dual boot the other language.