|
|
|
|
|
by alex201
790 days ago
|
|
I've never quite seen eye to eye with Daniel Dennett. His tendency to reduce the inexplicable to what he's confident he understands has always made me wonder if a challenging childhood might have fostered his distrust of the mysterious. Whenever I admire Nobel Prize laureates like Roger Penrose, who argue that consciousness isn't just software running on the brain's hardware, I can't help but feel a twinge of pity for Dennett and his like-minded peers. I can almost hear him reflecting, 'Wow, that was a wild ride, but boy, was I cranky! I wish I could have another go at it. |
|
Personally, I think starting with things are known to exist, which have a physical basis, is a great start, and untestable assumptions should be kept to a minimum. Just because it would be delightful to contemplate that ornately feathered technicolor quantum unicorns are actually underlying all of reality, it isn't productive to consider until there is a reason to.
Penrose is no doubt a genius of high order in his domain, but consciousness is not one of his domains. Saying consciousness is the result of quantum effects in microtubules explains nothing -- it is just a very tiny rug which one could imagine is hiding the truth, as all the larger scale hiding places have been inspected and found lacking.
You'd think that with the stunning (and mostly unexpected) success of LLMs would expose the fact that simple, soul-free, mechanistic computations can produce some really amazing capabilities. The human brain is orders of magnitude larger than GPT4, plus it has a wildly more complex architecture than today's neural networks. To me, it takes little imagination to see how everything could be explained in purely physical terms.