[tlb]

The last 80 years have had a string of successes for arranging simple calculators in ways that produce impressive results. Google search, world champion chess players, Fortnite are all examples.

I don't know of any impressive results with arranging forks. If there were, you could model the fork behavior on computers and probably run it 1000000x faster.

It's certainly possible that some other elements than the simple calculators we use today will lead to the big breakthrough in AI. Perhaps quantum computation is needed. But right now, arrangements of simple calculators seem like the most promising.

[goatlover]

> The last 80 years have had a string of successes for arranging simple calculators in ways that produce impressive results.

I view that as biological intelligence figuring out all sorts of clever ways to make calculating tools produce impressive results.

> Perhaps quantum computation is needed.

But why invoke physics if we're comparing biological intelligence to our computing tools? Isn't it enough to note the big differences? And here I'm not talking so much brain function as I am being a living, social animal that has to survive with a vast human culture.

This is the difference between looking at computers as as stepping stones to artificial replacements as opposed to enhancing our abilities. The AI stuff captures the imagination, promises fully automated utopias, scares us with apocalyptic scenarios, and is the stuff of lots of SciFi. While the reality is that computers have always been tools aiding human intelligence.

For some reason we view these tools as one day being like Pinocchio. The mythos of the movie AI is based on that vision where the robot boy becomes obsessed with a the blue fairy turning him into a real boy so his adoptive biological mother will love him and take him back.

But maybe like in the movie, someday the sentient robots will show up and gave our robot boy his wish.