There's No Fire Alarm for Artificial General Intelligence • Hacker News Board

Re: There's No Fire Alarm for Artificial General Intelligence

#111

Post by ianai » Sat, Oct 14, 2017, 2:59 PM UTC

mark_l_watson wrote: post #62

Great read, and I don’t mind at all that the last section was a pitch for donating to MIRI. I have been an AI practitioner since 1982 and have enjoyed almost constant exposure to people with more education and talent than myself so I feel like I have been on a 35 year continual learning process. I think that deep learning is overhyped, even though using Keras and TensorFlow is how I spend much of time everyday at wor…

Got any suggestions for a knowledge source of AI at a 100-1000 foot view? Ie not stuck in the weeds, but enough to know what’s going on nd where.

Re: There's No Fire Alarm for Artificial General Intelligence

#112

Post by notahacker » Sat, Oct 14, 2017, 3:07 PM UTC

marvin wrote: post #96

notahacker wrote: post #56

Earlier quoted context omitted.

Alphago is the equivalent a 1950s team coming up with a much better transistor that everyone expected to see in the next ten years in months AGI is a 1950s team surprising absolutely everybody by secretly spending a couple of years and a couple of billion dollars and creating an iPhone. By the time the iPhone was actually possible, everybody thought it was an impressive use of existing technologies rather than requir…

A better analogy than an iPhone, is the Manhattan Project. A phenomenon is discovered and then published: uranium can be split by neutrons, which liberates a relatively huge amount of energy. Some scientists hypothesize that a chain-reaction that produces neutrons to sustain itself can be produced with uranium, but see no practical path to accomplish this phenomenon. A couple of isolated groups of scientists then dis…

Is it really a better analogy? The Manhattan Project had straightforward success criteria (either the chain reaction sustains or it doesn't), was undertaken at enormous speed not least because scientists working on the project were convinced that even with limited information already in the public domain there was an imminent threat of the Germans building one first, and especially in those circumstances was unlikely to produce plausible useful byproducts before the success criteria was met. More of the budget was spent obtaining the fissile material the theory required through industrial-scale application of experimentally-known techniques than trying to figure out just what it was that made things go boom. The bomb test worked first time.

By contrast, AGI has been desired and actively and openly worked towards for decades and we haven't even decided what it is yet. And it's hard to imagine the intermediate non-threatening intelligences and their byproducts wouldn't be so impressive an advance that those working on them wouldn't be willing to share them. Without wishing to trivialise the immense amount of intellectual effort and industrial production that went into the Manhattan Project, sentience is a more complex goal than a chemical reaction. Or indeed an iPhone.

Re: There's No Fire Alarm for Artificial General Intelligence

#113

Post by icebraining » Sat, Oct 14, 2017, 3:10 PM UTC

namelost wrote: post #66

icebraining wrote: post #53

Earlier quoted context omitted.

I'm not saying alrs should. The decision to leave is reasonable. The decision to judge others ("What is wrong with people?") is not.

That depends whether the other occupants knew about the safety features and record of elevators. If they did, then they were making a sound judgement. If they didn't, then they were being completely reckless.

I guess it's possible the parent poster took a poll before leaving the elevator, but I have to say I find it unlikely.

Re: There's No Fire Alarm for Artificial General Intelligence

#114

Post by icebraining » Sat, Oct 14, 2017, 3:12 PM UTC

musage wrote: post #74

icebraining wrote: post #53

Earlier quoted context omitted.

I'm not saying alrs should. The decision to leave is reasonable. The decision to judge others ("What is wrong with people?") is not.

They're saying they think it was unreasonable to stay, you say it was reasonable to leave. Both are judgements about the decisions of others.

I don't think all judgements about the decisions of others are equivalent. Do you?

Re: There's No Fire Alarm for Artificial General Intelligence

#115

Post by icebraining » Sat, Oct 14, 2017, 3:22 PM UTC

musage wrote: post #79

icebraining wrote: post #59

Earlier quoted context omitted.

We currently have thousands, or even millions of intelligent entities (humans) which think pretty wild and dangerous things. We usually just tell them "shut up Bob and take your meds" and that's it. Sometimes we regrettably kill them. Why would such AGI have the means of turning all of the planet into anything? I mean, sure, I also think the Terminator is a decent movie, but that doesn't make it a reasonable blueprin…

> We usually just tell them "shut up Bob and take your meds" and that's it. Other times, Las Vegas and we shut up because we come up empty -- at best talking about gun control as if someone choking 3 people on a bus was substantially better and not still something where we come up empty. > Why would such AGI have the means of turning all of the planet into anything? How does getting off track with that address the ma…

It wasn't an offtrack; I was pointing out that we already have intelligent beings that don't share our goals. Sometimes it does have bad outcomes, but I don't remember it ever having "destroying all the planet, all the animals, all the humans, all the art". Which is why I'm asking why would adding another intelligent entity make us fear it.

Re: There's No Fire Alarm for Artificial General Intelligence

#116

Post by tyingq » Sat, Oct 14, 2017, 3:34 PM UTC

maxerickson wrote: post #84

And in the meantime businesses and governments are still going to deploy weaker AI to their own ends.

There is a lot of focus on strong ai. The dangers of weaker ai implementations, working together, or tied to dangerous things (nanotech, biotech, nukes, trucks, drones, etc) seem significant to me. Especially if you throw in things like an ad hoc ability to make new connections.

Re: There's No Fire Alarm for Artificial General Intelligence

#117

Post by Synaesthesia » Sat, Oct 14, 2017, 3:36 PM UTC

etiam wrote: post #93

As far as I'm concerned this whole discussion is severely hampered by failing to differentiate between intelligence and agency. Almost all of the bugaboo about runaway superhuman organisms comes down not to machines learning and reasoning about the world but to the effective high-level objective function controlling the actions of an autonomous system. Not making the distinction obscures important things. For one thi…

Yes. We can’t even define what so called “AGI” will even be. And we have still not solved many mysteries of human consciousness, or begun to touch on them.

I for one have been looking at the problem of ai’s playing Starcraft 2, and the decision making required, such as when you scout your opponent’s army choice, or tech, how to respond. So far they’re very far from solving that, but if progress is made, I’ll be impressed. That’s a very different kind of problem as say image recognition and classification. It requires planning. It’s a very difficult game even for humans to understand. Currently the autonomous systems can’t even play it.

Re: There's No Fire Alarm for Artificial General Intelligence

#118

Post by JabavuAdams » Sat, Oct 14, 2017, 3:49 PM UTC

deong wrote: post #106

sago wrote: post #33

I've been around AI since the end of the last big hype in the late 80s. The recent leap in machine learning has felt rather hyped to me. I don't think AGI is near. But I find myself agreeing with this article. Strongly. And I have long suspected that, we miss a lot of the significance and opportunities in AI, because we have only one exemplar of 'higher' intelligence: a human being. AI folk are so concerned with gett…

The article makes a lot of good points, but for me, the critical error is in assuming that if short term prediction is hard, long term prediction must be massively harder. He asked a panel for the least impressive thing they did not believe would be possible within a few years. In other words, pick the point closest to the boundary of that classifier. Obviously my future knowledge is imperfect, and anything close to…

> I don't think 30 years will see a machine that can make dinner by gathering ingredients from a supermarket, driving home, and preparing the meal.

Really? Why not? Once or twice, if we cherry-pick its performance, or reliably?

This is really surprising to me.

Re: There's No Fire Alarm for Artificial General Intelligence

#119

Post by JabavuAdams » Sat, Oct 14, 2017, 3:55 PM UTC

Santosh83 wrote: post #31

Why do we fear other intelligences anyway? Isn't that just a sign of our own immaturity? Maybe we need to evolve more before we start think about creating AGI...

The Dodos should have feared us. We don't even hate Dodos, but we killed them all anyway.

Re: There's No Fire Alarm for Artificial General Intelligence

#120

Post by johansch » Sat, Oct 14, 2017, 3:59 PM UTC

astdb wrote: post #28

Having done some work with the state-of-the-art of AI, I personally don't think AGI is near - might not even be possible. But the catch is the unreliability of (even expert) predictions on technology futures. My take is that it's worth taking pragmatic steps towards studying AI safety measures (i.e. OpenAI), but not going as far as to talk the likes of 'AI research regulation'.

> Having done some work with the state-of-the-art of AI, I personally don't think AGI is near - might not even be possible.

(Just venting here, not even primarily at you.)

360k babies are born each day. Clearly it is possible to reproduce intelligent machines. The only way it would be impossible to artificially do the same is literally if life was a magic, non-physical thing. I wish people who state things like this would also state any religious beliefs that lead them to think so.