Suppose for a moment that a neural network does in fact mimic how neurons work. (I am not sold on this, as it is based on 1960s neurology.)
If this is the case, then the most powerful learning computer we know of is the human brain, particularly in the first several years of life
-
-
And this learning computer trains, 24 hours a day, seven days a week. And after 5 years, it has the speech capabilities of a five-year-old.
-
And this learning computer is dramatically more complex than any we have been able to build. Ever. Without comparison.
-
And we expect our artificial mimics to come in, despite barely understanding *them*, train for far less time, developing far less ability...
-
Using algorithms that were basically invented 50 years ago
-
To come in and save humanity from its problems. Would you trust a five year old to negotiate a peace treaty?
-
Artificial neural nets, along with other similar machine learning techniques, are wonderful, amazing, useful tools. I love them!
-
But can we stop with the comparisons to a biology we do not understand? And the rosy visions of the future that whitewash their limitations?
-
Neural nets are not AI. They are not surrogate people. They cannot replace empathy. They follow gradients we can't picture, that's all.
- 4 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.