Thank you for the thoughful response, Dimitris!

G. Hinton is actually "suspicious" of both backpropagation and CNNs. He thinks we should get closer to the brain (the last quote). He isn't saying that today's systems don't work. But to go further we may need to rethink everything from scratch.

I think we should try to create AGI (although I'm not sure about the potential social and ethical consequences). However, I'm not in favor of progress as an end in itself.

About mimicking human biology, it's true that it hasn't been necessary historically. We have taken inspiration from Nature to advance technology many times. With the brain, it may happen similarly. However, we don't understand the underlying workings of the brain yet. It's not that the artificial system doesn't need to simulate the biological system. It's that we don't even have the scientific understanding of the latter yet.

It's funny because I'm an aerospace engineer so the example you used is familiar to me. We understood aerodynamics and flying mechanics before building planes (or better, we were understanding it at the same time). Even if planes don't resemble birds, the physical laws goberning both planes and birds were understood. That's not happening with neuroscience yet.

So my question is, how can we build something similar to the brain if we don't understand how it works?

AI losing its magic is something very subjective I think. It's good that it didn't lose it for you!

Again, thank you for the response! Appreciate a lot your last paragraph! It's good to have a healthy debate! Cheers :)

Writing at the intersection of AI, philosophy, and the cognitive sciences | | Get my articles for free: