Slashdot Mirror
Elon Musk Is Paying For Free Streaming of a New Documentary about AI Dangers (syfy.com)
An anonymous reader quotes Syfy.com:
There's a new documentary warning about the perils of artificial intelligence out there, and Elon Musk wants you to see it. So much so that he's making it available to stream for free this weekend. The documentary -- Do You Trust This Computer? -- explores the rise of machine intelligence and its possible consequences... Check out the trailer, and then proceed to be creeped way the hell out.... "It's a subject that I feel we should be paying close attention to," said Musk in a news release. "I think it's important that a lot people see this movie, so I'm paying for it to be seen to the world for free this weekend."
Musk attended the premier of the film with the creator of HBO's Westworld, and tweeted Saturday that the video had 5 million views in just 36 hours.
Musk himself is interviewed in the film, warning of the dire possibility of "an immortal dictator from which we can never escape."
Musk attended the premier of the film with the creator of HBO's Westworld, and tweeted Saturday that the video had 5 million views in just 36 hours.
Musk himself is interviewed in the film, warning of the dire possibility of "an immortal dictator from which we can never escape."
Can anyone ever tell me of a documentary you had to pay to see? How is this news?
https://xkcd.com/1968/
Also isn't "this weekend" referred in TFS the wrong one for the vast majority (if not the entirety) of the world?
Just because it's alarmist doesn't mean the film is wrong. I'm most worried about the fact that the creators of neural networks often don't understand how they are operating; as in, why it has the inferences it does. I wonder if it's possible to train a neural net, then iteratively reduce its complexity without affecting its performance, down to the point where we can understand its operation.
Corruption is convincing someone that the selfless ideal is the same as their selfish ideal.
I've watched this to about the nine minute mark, at which point it froze, waiting, I presume for the stream to continue. So far, I've seen a lot of woo and the usual modern glitter; rapid montages of computer graphics, fish-eye lens shots of people stroking touchscreens, and infrequent and REALLY ANNOYING deliberate "glitches" where the screen flickers and distorts with that squeaky Hollywood "computer video malfunction" sound. A couple of respected people in the field offering their opinions; some vox pops of kids and just good old every day folk admitting they use computers a lot, and for Christ's sake, that shot from "Terminator 2:Judgement Day" where the T-800 crushes a skull with its metal foot.
The bias is obvious, and I'll keep watching as soon as the buffer opens again, but I'll be surprised if anyone in this comes even close to a solution to the problem it's posing.
(Incidentally, getting really tired of hearing pundits say "We're making advances in ageing studies which means we could be immortal soon." What you mean "we", white man?)
Knock out all electronics on earth... we should be developing that first :)
[($)]
Personally I'd rather be ruled over by an immortal AI dictator a la The Culture, than a hypocritical moralizing human (or group of such humans). If anything, it would be resistant to bribery and appeals to its ego.
Corruption is convincing someone that the selfless ideal is the same as their selfish ideal.
AI does not exist and probably never will. What we have right now isn't AI, it's not conscious, it's not learning. It's just software that stores large amounts of data. It has no intelligence or creativity or emotion to act upon anything.
This film is equal amounts of FUD and fearmongering. Almost nothing else (except some jobs will be replaced by AI - but that's been happening for over 200 years, except I'd replace "AI" with technology).
Most hardcore AI experts (and Musk is not one of them) don't see AGI happening any time soon. We just have no idea what intelligence and consciousness are. Not a freaking clue.
Back propagate the same error correction to all nodes makes time series prediction notoriously inaccurate. Feeding that error forward only increases the noise.
The "deep learning" currently done by AI researchers, and the type of "strong-AI" in the Terminator, really have little to do with each other. Hollywoodesque strong-AI is still science fiction, and will be for a while.
Here's some good advice about what to worry about: https://xkcd.com/1968
So, assuming that they are right, and developing AI is going to end up in a super-intelligent conscience that enslaves humankind. OK. Now, what's the counter-strategy? Forbidding AI? Sure, tell me that today's powers will stop developing smart weapons, trusting that the other powers will stop too.
It's a similar situation to global warming. You cannot do much about it, people's minds are too short-term for that. Most people will choose some money now rather than double that quantity in a year. History will have to run its course based on the capabilities and constraints of today, not on any conscious decisions that we could make.
Perhaps we are just a stepping stone to a higher organism, a big mind, dependent on millions of little ones to keep it alive. Hey! that's just what our brains are! Perhaps at some time in the past, cells should have decided that they didn't want to be slaves of a consciousness themselves made (in both senses). If they did, it didn't work too well in the end. Well, in my humble opinion, we have about the same chances of altering whatever course is due now. So, what's the point of running hither and tither like a headless chicken worrying about it. They could as well have made a documentary about how we are all going to die, for all that's going to be of any help.
Rome taught me patience and assiduous application to detail. Virtues which temper the boldness of great, general views.
I'm watching it right now.
I don't trust atoms -- they make up stuff.
I was entertained.
"First they came for the slanderers and i said nothing."
FTFY
#DeleteFacebook
I'm afraid of people abusing the english language.
#DeleteFacebook
Maybe we'll be glad about Google, Facebook and others creating so many web tracking cookies in a few years.
#DeleteFacebook
I'm concerned about ALI - arrogant lack of intelligence. ALI in Teslas which kill their drivers. And ALI in Musk, who was responsible for cars that killed drivers, and whose company waited for its first fatality before issuing an update to slow cars to a stop when the driver is not paying attention (or is incapacitated).
+1 funny
#DeleteFacebook
Could anyone find that indeed Mr. Musk is paying for it on the documentary's site?
"Sum Ergo Cogito"
I just watched it (Arizona). The deadline must be Sunday midnight.
Cars need not be self-aware, so we don't have to wait for general AI. In fact, self-aware BMWs might be the apocalyptic weapons this this account was talking about.
Musk has (Ivy League) degrees in physics and economics. Gates, and I'm sure Hawking, have a deep understanding of what computers are (computers are devices that execute binary instructions. Also, they allow users to manipulate information encoded in those binary instructions). Gates built a compiler.
A device does not need to be conscious "like us" to be able to operate in the real world. It doesn't need to have physical systems like us. It just needs to process information and change its environment.
"Information processing" and "affecting the environments" are the keys here. It doesn't need to be "like us."
You can look at DNA as an information-archive. An inefficient, time-consuming gathering of information built over the eons. The forces driving the creation of DNA created humans. Humans are the only animal which can store information outside of itself. One of the common denominators of life is that it (net) increases entropy. Another is that it (slowly or quickly) gathers information, in DNA or otherwise. What if that information gathering is somehow a driving force in the universe?
Humans have a deeply-seated, deeply held core conceit: that we are separate and above the universe and nature. That's why we have the terms "man-made" or "artificial" versus "natural." But humans are a product of the universe as surely as beavers and birds are. And we don't look at their nests or dams as anything other than being natural - products of nature. Thus, human creations are also "natural" - but look at how the language holds back our thinking. The definition of "natural" is "made by nature" and the definition of "artificial" is made by man. There is no word to describe man's creations as being a product of nature, or natural in anyway. And this gap is an obstacle in understanding the universe and man's role in it.
So: tl;dr: Computers are dumb electronic machines totally unlike us. Doesn't matter. All they need to do is process information and change the environment autonomously to present a threat.
And IF information processing is a core drive in the universe, and man was an accelerated information gatherer and processor, and man - a product of the underlying forces in the universe - created a next-level information gatherer and processor... would that creation then obviate man's role in information gathering and processing? I mean, carbon-based life would still do a reasonable job of creating net entropy. Maybe the underlying forces in the universe will still create a need for carbon-based life.
Maybe creating entropy and gathering and processing information is part of some game the universe is playing or being used to play. And we're in the process of birthing the next generation information gatherer and processor. Let's hope the mother doesn't die in the birth.
Actually, at run-time, we don't know how they're working. We do know how they're working at compile-time.
I saw these snippets in an article recently in The Economist (probably paywalled):
"The reason for this fear is that deep-learning programs do their learning by rearranging their digital innards in response to patterns they spot in the data they are digesting. Specifically, they emulate the way neuroscientists think that real brains learn things, by changing within themselves the strengths of the connections between bits of computer code that are designed to behave like neurons. This means that even the designer of a neural network cannot know, once that network has been trained, exactly how it is doing what it does. Permitting such agents to run critical infrastructure or to make medical decisions therefore means trusting people’s lives to pieces of equipment whose operation no one truly understands.
If, however, AI agents could somehow explain why they did what they did, trust would increase and those agents would become more useful. And if things were to go wrong, an agent’s own explanation of its actions would make the subsequent inquiry far easier. [...]
One of the first formal research programs to attempt to crack open the AI “black box” is the Explainable AI (XAI) project, which is being run by the Defence Advanced Research Projects Agency (DARPA) [...]
The program does this by drawing on the assistance of a second neural network which has been trained to match the internal features of the agent doing the recognising (ie, the pattern of connections between its “neurons”) with sentences that people have written, describing what they see in a picture being examined. So, as one AI system learns to classify birds, the other learns simultaneously to classify the behaviour of the first system, in order to explain how that system has reached its decisions. [...]"
Separate note: In Neuromancer, the AI's come in pairs, Wintermute and Neuromancer. It would be interesting indeed if AI's going forward do come in pairs. Wouldn't be the first intelligence to operate in pairs.
One more thing: Gates and Musk understand systems. At a human level of course, but a level much higher than most humans.
Thankfully I'll be dead before it gets too bad.
What's the difference between our computers being controlled by some fictionally competent AI vs. our computers being controlled by Microsoft, Apple, and Google? They're all immortal (corporations are people, remember?), not democratically accountable, and don't give a sh#t about you and me. The only difference I can see is that corporations have governments bending over backwards to ensure their success. In this respect, AI would at least meet some significant resistance.
Debate is a form of harassment. Do not question my truth.
What I don't trust is the programmer who decides what to do with the computer or the management who decides what the programmers should do. The end result will always be the most profitable lowest common denominator trash. Hardware has always been my passion, whether auto (mechanical) or computer it ALWAYS does what you tell it to do, even it is not the right thing.
errr....umm...*whooosh* *whoosh* Is this thing on ?
There's a lot to be concerned about here, but the thing that everyone seems to miss, over and over, is the fact that we can't secure our computers against humans, let alone an AI with infinite patience. A few years ago, all of the 128 page security clearance applications for the entire United States were digitized, and online.... who was stupid enough to let this happen? Everyone was surprised and shocked when it happened, but I bet most of you don't even remember it any more.
All this data is eventually accessible via the internet, and there's shit for security protecting it. One lucky rogue human is all it takes to take the whole thing down. I'd be deeply surprised if someone, somewhere, isn't training an AI to take over compute resources.... and once that gets sufficiently good, it's game over, because nothing is secure.
It's possible to radically increase security, and do it in a user friendly manner... but this requires re-writing everything based on a new security model. (The principle of least privilege), so it's not a "magic bullet", but rather an expensive one.
I hope we decide to spend the resources and fix security... but it's a faint hope.
My wife is addicted to the grade-C documentaries one can watch on Netflix or Youtube. Thankfully she uses a headphone. I glance over at her tablet from time to time and note that very often she is watching a 'video' that is a camera panning over still images, then a cut to a talking head with obligatory bookcase in the background.
It's grown to be very cheap to produce low grade 'documentaries' with modern cameras and digital video editing equipment.
Not sure if I should type *whoosh* or not, which is bewildering.
Yes, but AM was driven by pure sheer emotional hatred of humans. Almost the opposite of what one would think an 'evil AI' would evolve as a central motivating core. Evil AI would be detached and cruelly inhuman, not motivated by hatred.
Ellison is a brilliant writer, but I don't credit him with a good understanding of computers and Artificial Intelligence.
The self-aware BMWs would quickly be obliterated by the self-aware F150s. You can count on that.
You don't need the fucking Terminator for AI to be used for evil.
How about China? They've got that fancy new Social Credit rating system. Bet that's pretty hard to maintain, there's a lot of data to sift through. So they build an "AI" to do some of it for them. AI now controls a significant portion of peoples lives and it doesn't have access to a single nuke!
Even better, whenever the AI makes mistakes and ruins peoples lives, there's no one human to blame! The AI told us to do it, not our fault! We fixed the bug and of course there aren't any more, ha ha that's preposterous. Then there's the fact that we're already getting results from certain systems that we don't quite understand how the system actually got to the results, or at least not without a significant amount of reverse engineering. Humans already have a hard time with accountability, why not build a society with none whatsoever?
I'm less worried about AI going Skynet than I am worried about it being the most efficient system of social control in human history. That, and that the people who are developing it are going to do so recklessly, just like every tech has been. Does anyone really trust China, the US, or the Silicon Valley Data Barons to develop AI systems that actually help people? By "help" I obviously don't mean "help people give us more data", so that's right out.
So it doesn't matter if the NSA uses AI systems to more efficiently process all that juicy data they're stashing away in those giant data centers? *yawn*
Corporations knowing, in absolute terms, more about you than you know about yourself? *shrug*
As long the AI isn't nuking cities and gunning down kids in the street it's no problem whatsoever. Got it.
My wife is addicted to the grade-C documentaries one can watch on Netflix or Youtube.
One can also watch grade-A docos on Youtube and Netflix so maybe it says more about your wife than the content provider.
It will be free until the offere expired yesterday. hmm. I see grammatical issues, and others too..
Beware the man bearing gifts, especially one's you actually have to pay for.
Besides, fears over the singularity are not that insightful. Nobody knows what will happen, so its easy to describe fearsome possibilities. In my opinion AI will be controlled by capital, and will benefit those who have the capital at the expense of those that do not, just as it has been for a long time. Why would people be wary of that future when it has been readily accepted as reality for generations already? I think our current administration proves my point.
Lots of comments, but almost no one watched the video. At least this comment watched the video.
I did, and didn't like it. I am VERY sympathetic to the view that AI/ML can be dangerous/unethical, but this documentary is not well made. It has high production values, but this video plays scary music while mocking fringe characters, to get you scared of AI in general, and 5 minute montages of computer stuff with different scary music. If it dialed itself up a little bit more, it would be a good parody of itself.
Unfortunately, it's not completely terrible, it does mention some real things to be concerned about (truck drivers losing their jobs, big data means big surveillance, etc.). And then it shows some creepy uncanny valley female robot doll with some weirdo guy who invented it saying some BS, and I think we're supposed to just be creeped out. But it offers NO solutions/suggestions, it's just meant to be scary in a general way. I don't know what it wants to achieve--do we need to destroy all computers?
Let me use ML (machine learning) as a shorthand for all varieties of neural nets, since that's how non-AI people refer to it, and AI is more general, including general learning capabilities. ML exists today, and in its simplest form, it's complex pattern matching. Show a neural network a million pictures of sheep, and it learns to pick sheep out of pictures. The documentary said ML was something different, but it's wrong.
Problems with ML/neural nets: requires massive data (surveillance issues); bias in collecting data is passed through to bias in learning (google "prank a neural net" to see how an ML net that's excellent at picking out sheep in photos, when given a picture of a sheep indoors makes the ML think the sheep is now a dog, and other even funnier mishaps, since all the training photos of sheep were taken outside); still very much an "art" and not a "science" in applying it to real problems; much current ML is done in the cloud (Siri, Alexa, Google Home), which creates more surveillance; and the creepy guys on the fringe of ML (like Cambridge Analytica). Almost none of this was in the documentary.
Solutions: true data privacy, make it illegal to sell user information for any marketing purpose; make clear rules on liability for ML giving bad answers (self-driving cars; loan processing; surgery); etc.
For AI, the documentary made this point quickly, then moved on: General AI may be just 3 breakthroughs away, which may mean 20-30 years. People who say "we don't know how to build this now, don't worry!" are just distracting from the real issue. Let's not start with fear, let's start from another direction. Here's one example: would it be murder to turn off a general learning AI? What rights should a learning intelligent AI have? Doesn't it make sense to have some ethical rules for investigating a truly general AI, at least to avoid doing unethical things TO the AI itself? Note that these rules don't apply to basic ML neural nets, since they are just pattern matching. And similarly, rules to prevent a Morris Worm virus-like effect, where something gets out of hand unintentionally. Not that the main fear is Terminator wiping out all humans, but all self-replicating systems need constraints put on them somehow, or they will run into constraints on their own.
If the AI community cannot create some basic ethics rules, then expect more documentaries, blurring the difference between ML (exists today, being used today, ethical issues almost all involve human-level problems) from general AI (doesn't exist, but might in a few decades, project any fears you want onto it) like this one which scare people into doing something no one will be happy with.