With current AI algorithms, it will not be possible to reach the intelligence-per-watt efficiency of the human brain, because that efficiency is the product of a vastly long biological evolutionary process encoding multiple diverse learning mechanisms (in-context learning, reinforcement learning, mental simulation, Bayesian inference) beyond what next-token prediction alone can produce.

97:21transform based networks and also like our type of let's say the the current architectures that are landscape of architectures that are available. They gave us with scale they gave us in context learning capability. The thing that actually emerged from next token you see the the word is important emerged from next token prediction. Intelligence for me is an emergent property. If you want to get into like if you want to miniaturaturize kind of

97:46intelligence bring it like to the to the physical world. I don't believe with the current set of algorithms like we would be able to get close to the let's say intelligence per watt that human brain is actually like providing right you're not going to get there why because I believe human brain over the we also have to like consider the amount of energy that went into design of humans

98:09you know as a whole biological evolution is actually a very very long kind of process and a lot of the pre-training process that people say oh a lot of you know like like you're going to see the entire world. Humans do not need to see the entire internet to be able to reason about something, you know. No, but humans have have gone through like years of evolution, you know. So, I would

98:27actually attribute it a lot to the evolutionary kind of aspect of things. But again, what I would say human intelligence came with multiple mechanisms of incontext learning. you know you just don't have when when when when when our current AI systems do in context learning they learned a vague representation of one algorithm which is list square you know it's basically list the square you know so what they what

98:51they what they figured out is basically gradient descent in a mushy way you know like for some use cases with some examples you would be able to make the system understand and give you like kind of the next example these are kind of beautiful property this is an intellig this is what I would call in emergent property of these systems. You set the algorithm to be next token prediction.

99:10You got a grad a vague version of a gradient descent in context learning capabilities for humans. You just you don't have only emergent gradient. You can learn by examples but you can also do reinforcement learning. You can run simulations in your head. You can do all sort of algorithms. You can do bijian statistics in your brain. You know you can you you see what I'm saying? So you