People Actually Prefer Taking Orders from Robots
Aug 26, 2014
Here’s something new – people actually prefer being led by AI, according to an MIT study. Who would have guessed? After all those Hollywood horror stories and all the scoffing at mindless unemotional robots, it turns out we actually like artificial intelligence.
The study involved complex assembly on a factory floor. When the human workers were instructed by a robot, they “reported feeling at their most efficient and effective.” It’s interesting that the human workers weren’t necessarily more efficient, but they did feel more efficient. That’s the cool thing, because we’ve always known machines were efficient. Only now are they saying that we like machines, too.
Well, what’s the explanation? The article was disappointingly silent on why people might be happier with an AI boss. So let’s brainstorm a bit. Here are three possible explanations, and for me the third seems particularly interesting.
Six Ways to Recognize a Crackpot AI Theorist
Aug 19, 2014
Why haven’t we achieved general artificial intelligence yet? One reason is because the field of AI is full of crackpots! Too many weirdos with their crazy superstitious ideas and get-rich-quick schemes.
And they’ve scared off the smart people. At least that’s what Nick Bostrom seems to think. In an interview he said, “A lot of academics were wary of entering a field where there were a lot of crackpots or crazies. The crackpot factor deterred a lot of people for a long time.”
I suppose all honest, self-respecting AI theorists need to think about whether they themselves might be crackpots. So here goes. I won’t try to tell you I’m not a crackpot, but I’ll just give you a few suggestions for how you can decide for yourself. Here’s a quick list of six ways to recognize a crackpot.
Let’s Go Ahead and be ‘Biological Boot Loaders’ for AI
Aug 11, 2014
Elon Musk recently tweeted: “Hope we’re not just the biological boot loader for digital superintelligence.” But I just said, “What’s the problem with that?” I think it’s great for humanity to be the “biological boot loader” for AI. I love that expression, in fact. It’s really perfect, and it sums up where we stand in the whole evolutionary scheme of things.
Yes, let’s imagine human beings paving the way for AI. Let’s say evolution just gets started in the biological sphere with DNA, but then it jumps to the silicon sphere for even more evolution of machine minds. What’s so bad about that?
Nobody lives forever, despite what Ray Kurzweil would have you believe. No species can carry on forever without evolving into various new species. Does Elon Musk want to bring evolution to a halt just because he has managed to be alive at this moment? Sounds kind of selfish. He ought to loosen up.
Imagine the last Homo Erectus people – how would they feel if they knew about us modern humans? What would they think if they could foresee all the great cultural, artistic and technological wonders we would create? I think such a visionary Homo Erectus person would be awfully proud of us. The Homo Erectus people could think of us as their children. And don’t we all want our children to be a success, to surpass our own achievements?
Look at the Huge Difference between Artificial and Natural Selection
Aug 5, 2014
I’m always saying that we can’t program a true AI, but it must evolve by natural selection. That’s the only way the AI can really be conscious, alive and morally equivalent to a human being.
The problem with this, which various people have pointed out, is that natural selection is out of our control. Something that evolved by natural selection would just be pursuing its own self-interest, and it could even compete with humans. That could get dangerous! An intelligent machine that evolved without human intervention or any kind of tweaking might be “unfriendly AI.” It might declare war and wipe us out.
In short, if something evolved by natural selection, it would have a mind of its own. And that of course is the whole point.
Let’s pause a moment, though, and just consider the difference between natural selection and artificial selection.
Why is Star Trek so Bad at AI?
Jul 28, 2014
I re-watched the Star Trek episode called “Datalore,” which is from the first season of The Next Generation, and this is the one where there are actually two androids – not just Data, the regular crew-member, but also Lore, who is Data’s evil twin. It made me wonder again why Star Trek is so dumb in its portrayal of artificial intelligence.
Well, first of all what’s wrong with the AI in Star Trek? I think Data is fine as far as he goes, but the main problem is that nobody in the Star Trek universe understands how he works. Even Data himself does not seem to know how he works. Only the mysterious Dr. Noonian Soong knows, but why didn’t he tell anyone else? He could at least have written it down somewhere. This is maddening.
Why are there only one or two android models like Data instead of thousands or millions? There should be whole factories building androids like Data. Considering that Data is more or less immortal, they should leave all the dangerous exploring and stuff to androids like Data, instead of making fragile humans do that risky work.
OK, and another ridiculous thing is the way Data always wants to become human. This is just blatant anthropomorphism on the part of Star Trek writers. Obviously Data is at least as good as a human being, and there’s no reason why he should think less of himself for being a machine. I can understand Data wanting to learn things from humans, but he shouldn’t want to be a human. What’s with all this AI self-hatred?
Lots more posts -->