The AI on Your Nightstand


Image credit: “iPhone” by Philip Brookes on flickr.

 

Artificial intelligence is real. It is evolving right now.

The last post started to discuss the subject of artificial intelligence. I would like to emphasize that AI is all over the place. Your smartphone is bursting with it. It is used in medical diagnoses, movie recommendations, and speech recognition. The Queens College computer science department offers an artificial intelligence course, right alongside computer graphics and software engineering.

In other words, a form of AI is already here. Much of what we call intelligence already exists in computers. This is a reality we have to deal with.

Should we be worried?

At the end of last post, I linked to an article from Tim Urban at Wait But Why.

Edge1

If you’re even a tiny bit interested in the future of artificial intelligence– and who isn’t?– you should read this. It’s a funny, thorough, and well-researched take on the topic.

The scary part is that Tim Urban thinks AI will singlehandedly determine the future of the human race. I think to an extent, he’s probably right.

Train4

What do you think? Do you agree with the article? Let me know in the comments.

5 thoughts on “The AI on Your Nightstand

  1. I don’t understand the notion of super-intelligence. Wouldn’t something that’s more intelligent than humans appear unintelligent? In which case, how is it intelligent?

    For instance, if it’s so smart that it can see that 2+2=2, but it takes a level of intelligence beyond our understanding to grasp the proof or grasp an algorithm which would produce the proof, then how is that smart? 2+2=4. Duh.

    And if it can produce a proof that we understand, or an algorithm we understand which could produce the proof, then someone else could have thought of it (and maybe did but just never got published because journals are more interested in publishing what robots write than what humans do, it seems), so it’s not super intelligent.

    On a different note, I’ve noticed that in conversations about AI, the notion of hackability is left out (I’ve never seen it mentioned). AI, whether ANI or AGI or ASI or A.*I, is hackable. It’s a thing. Like a computer, a network, your shoes, any man made thing, it can be made to operate in ways unintended by its designer. This is kind of a side-channel of the argument, so it doesn’t address any points, but it does, when taken into account change some of the assumptions.

    • I hear what you’re saying about hacking. That’s interesting. Is there an analogy in humans, do you think? Like, electronically stimulating people’s brains or nerves or something

      Also, it seems that you are working from the premise that if something does not appear intelligent to humans, it is not intelligent. Can you explain that premise?

      It seems like what you’re saying would apply if you instead looked at everything from a cat’s perspective. A cat doesn’t understand thermodynamics, so a person who understands thermodynamics won’t appear especially intelligent to a cat. The stuff we can invent with our understanding of thermodynamics might affect the cat’s life, though.

    • right, but the word only has the force of the meaning we’re capable of giving it. like if some dude were to tell you that he figured out that 2+2=4 and he couldn’t prove it in any way you could understand, and then he wore tinfoil clothes and asked you for money for reasons you don’t understand, or just started doing bizarre stuff for reasons you don’t understand, all on the claim that he’s transcended intelligence, what would you think?

    • 2+2=2, i mean. and about the brain, its hard to say that something that wasn’t designed by people has been hacked, because how do you know that that’s not part of its intended function? you need to be able to understand the mind of the designer in order to be able to determine if a thing’s being subverted

    • Do you really think there’s no such thing as evidence for intelligence? The guy on the street isn’t doing anything that is supporting his claim of intelligence. In addition, it is more likely that a human is crazy than exceptionally intelligent.

      But if something was designed in a sensible way, and we could observe it progressively increase in intelligence, why would we suddenly stop believing it once it passed our particular level of intelligence?

      Also, are you defining intelligence as an objective phenomenon, or as something else? It seems like you’re treating it as a label that humans give to things, but maybe not an actual reality. How would you define intelligence?

Comments are closed.