I hear you - its just that I dont think that it neccesarily follows that an AGI in whatever form will necesarily be unconstrained or free to improve itself. I also happen to think that the biological systems are in-fact…
I disagree with your argument, especially for point 1: these systems are massively constrained. The hardware they run on is fragile requiring massive amounts of power and tightly controlled environments. They don't have…
I think there may be a significant advantage to running your own server: you can controll how long you want to keep the data. For example, you can keep data from cameras recording the street/yard for longer and delete…
I remember reading a popular science article a while back: apparently we have managed to construct the complete neural connectome of C. Elegans (a flatworm) some years ago and scientist were optimistic that we would be…
I feel quite strongly that there is a large difference between Stable Diffusion and Copilot: with the size of the training set vs the number of parameters, it should be very difficult if not impossible for Stable…
Presumably you can use the 97% that is correctly transcribed to rapidly filter out the relevant content. This is likely to be only a small portion of the total content. Then you check 100% of that.
It seems that backpropagation might be a good abstraction for biological learing rules after all: "local predictive coding converges asymptotically (and in practice rapidly) to exact backprop gradients on arbitrary…
I hear you - its just that I dont think that it neccesarily follows that an AGI in whatever form will necesarily be unconstrained or free to improve itself. I also happen to think that the biological systems are in-fact…
I disagree with your argument, especially for point 1: these systems are massively constrained. The hardware they run on is fragile requiring massive amounts of power and tightly controlled environments. They don't have…
I think there may be a significant advantage to running your own server: you can controll how long you want to keep the data. For example, you can keep data from cameras recording the street/yard for longer and delete…
I remember reading a popular science article a while back: apparently we have managed to construct the complete neural connectome of C. Elegans (a flatworm) some years ago and scientist were optimistic that we would be…
I feel quite strongly that there is a large difference between Stable Diffusion and Copilot: with the size of the training set vs the number of parameters, it should be very difficult if not impossible for Stable…
Presumably you can use the 97% that is correctly transcribed to rapidly filter out the relevant content. This is likely to be only a small portion of the total content. Then you check 100% of that.
It seems that backpropagation might be a good abstraction for biological learing rules after all: "local predictive coding converges asymptotically (and in practice rapidly) to exact backprop gradients on arbitrary…