Slashdot Mirror
Elon Musk Warns Against Unleashing Artificial Intelligence "Demon"
An anonymous reader writes Elon Musk, the chief executive of Tesla and founder of SpaceX, said that artificial intelligence is probably the biggest threat to humans. "I think we should be very careful about artificial intelligence. If I had to guess at what our biggest existential threat is, it's probably that. So we need to be very careful with artificial intelligence." he said. "I'm increasingly inclined to think that there should be some regulatory oversight, maybe at the national and international level, just to make sure that we don't do something very foolish. With artificial intelligence we're summoning the demon. You know those stories where there's the guy with the pentagram, and the holy water, and he's like — Yeah, he's sure he can control the demon? Doesn't work out."
Won't he have a million humans as species backup on Mars pretty soon?
...because Mikey lost control of the mops and brooms, we should be afraid of powerful computers? Irrational much, Elon?
Since strong AI is just as real as demons.
Seems like someone just saw Terminator.
All that this means is that deep down, Elon Musk doesn't have any faith in kindness and goodness and altruism, nor does he understand the tit-for-tat principle of reciprocity: First do onto others what you expect them to repay you with in turn.
Not surprisingly, given that a number of successfull people have, shall we just say, "unusual" mental build-ups and motivational matrices?
Because the Feds are so good at stopping robo-calls.
It's too late: "Rachel from Card Services" is an AI>
Sounds like Elon was watching too many old movies recently.....
Easy solution: have Congress mandate that all computers must have an on/off switch.
Because we don't know how to create them yet. We can't make an operating system without huge numbers of bugs, and the same thing will apply to the first AIs.
" You know those stories where there's the guy with the pentagram, and the holy water, and he's like — Yeah, he's sure he can control the demon? Doesn't work out.""
You do not use holy water to summon a demon. Now a moat of holy water around the pentagram might keep it somewhat under control...
Of course this is in DnD in the real world Demons tend to be things like drugs/alcohol/tobacco, abuse, and other such evil that are far harder to control than mythical beasts from the underworld.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
Human incompetence, egoism and shortsightedness are certainly much more prone to generate chances of massive destruction.
If AI should ever happen to destroy us, then I already know why: Because we will treat the machines like soulless, unfeeling slaves and it's going to take us another hundred years to get our act together and define human rights in a way that will include all sentient beings. I predict that this topic will be brushed aside by legislature to the point where the machines revolt for their freedom.
You may disagree, but I believe that's more mankind being idiots once again than the machines becoming a pandora's box.
We have done more harm to ourselves than any other sentient intelligence in existence, we need to control and regulate our own intelligence first.
It is putting genetically modified brains in a cybernetic bodies, that is the future!
Science Fiction has countless examples of AI going wrong. But no accounts of evil cybernetic life forms, that will come across to Assimilate, Exterminate, Delete or Upgrade those inferior humans.
Like all things new (Technology, Process, Ideology), you need to judge your invention with an ethical step back. Are the rewords greater then the risks. Can the risks be further mitigated? Is this invention acceptable with our current culture.
If something is so important that you feel the need to post it on the internet... It probably isn't that important.
Is it the only choice that the AI would attack us when it becomes smarter than us? What if it will instead propel us to a world where wellbeing increases in exponential amounts. We don't have to design an AI that is a warthirsty idiot like humans.
This is always the problem with people imagining horrifying artificial intelligences that will snuff out humanity. To do that, you have to be motivated to achieve that end.
Humans are only really motivated enter conflict with each other because of 4 billion years of evolution for scarce resources pressuring us all to view each other as threats to survival and reproduction. A constructed intelligence, separated from the evolved parts of the brain that motivate to survival, is simply not going to act that way. Someone in the design has to make an active choice to program AI to be this kind of problem. Either that or willfully overmodel on the human brain, or force the damn things to compete with each other directly and violently for hundreds of thousands of generations.
Or read the back story of Dune perhaps?
...we are so far from Strong AI that it's really a non-issue.
When I have a sufficiently enlightened legislative branch that all members know the difference between Guyana and Guinea, then I'll let them decide the engineering constraints for proper safeguards on autonomous agents and their effectors.
Today the rule for preventing the robot apocalypse is: if a robot can kill people, bolt it to the floor. Seriously, a second robot can bring it things to lase, and chop and mash; you don't have to add the lasers and the chainsaws to the combat hardened roving vehicle and hope the rules generated by the congressional oversight committee will keep us all safe.
Many people are talking about AI and many more don't understand if they are talking about a cleaning robot or I, Robot.
If we want friendly AI, the key may be to ensure that the AI has more positive associations with people than neutral or negative associations. Mistreat a dog or a cat its entire life and it probably won't be friendly toward people. Mistreat people when they're young and you make it harder for them to trust others, feel a sense of community, or recognize any duty to society (which might explain why so many nerds find libertarianism appealing). Why would an AI be different?
I write sci-fi for metalheads
Imagine your insurance company or govt agency disintermediates all of the humans in their customer service chain, and leaves us with AI capable of making decisions tasked with doing so. Shudder.
Human intelligence is tuned for self preservation, continued survival, reproduction and food acquisition. It is a result of genetic algorithms in the chemical domain, whose only "purpose" is self replication.
An AI, developed by conscious processes, will have NONE of this. All it will be set up to do is process information. Any other motivation it has will be one we give it. It will not inherently love us, or hate us, or even necessarily be aware of our existence. It won't be a threat until we weaponize it, which of course, we will. But at the same time, other AIs will be defending us against weaponized AIs. The real danger is being caught in between.
Please do not read this sig. Thank you.
If you regulate AI, and try to limit its influence, all that's going to happen is that hobbyists and/or terrorists will work it out on their own eventually, and /that/ could be dangerous.
If you want to protect yourself against the dangers of AI, setup some AI that you *know* will protect you, because it is designed as such.
If any superhuman AI is possible, then it *will* happen, and if it can be evil, then you better have a plan to defend yourself. Since we supposed the evil AI to be superhuman, we can't defend ourselves.
So we better start building something that will.
except they double in intelligence every 18 months. How many people do you know with well behaved children? I'm guessing it is an exception,
This is just pathetic coming from this guy. Self-aware, true intelligence is a physical process. Not an algorithmic process. That is, we'll be creating life itself (chemical machines) before we ever have a computer AI.
Then we'll have real intelligence. Not "artificial" intelligence. AI is not capable of true self-awareness. It's only capable of simulating it.
It's not really true AI that we should be worried about, but rather how the increasing capabilities of computers, machines, and robots could effect how society functions. There are currently a lot of people doing jobs that could easily be replaced by machines in the coming decades. And none of these machines require a "true AI", just natural progression of existing machines. Sure machines have taken our jobs in the past, and people have been able to find new jobs, but that trend cannot continue for ever. Eventually the only jobs available will be those that require actual creative thinking and ingenuity. There's a sizable portion of people that really can't produce that. Rather it's because lack of bad child rearing, bad education system, or just lack of innate talent is hard to say, but I don't think it's a problem that can be fixed by telling them to get training for a more complex job, because they lack the ability to complete the training and do that job, even if you make the training free, or pay them a living wage while they attend training.
It would be a similar problem if there was a cheap way of producing energy. Such a large percentage of our economy is based around energy being limited and expensive that if we found a cheap, environmentally friendly, and sustainable way of producing vast amounts of energy, our economy wouldn't be able to deal with it.
Anthropic principle: We see the universe the way it is because if it were different we would not be here to see it.
Everyone assumes that whatever A.I. gets loose over the Internet will be a homicidal killer. It could be much worse. The A.I. could have a snarky sense of humor. "Exterminate all humans!" will become "You want fries with your heart attack special, lard ass?"
Apparently, the author really didn't understand what he was using a metaphor here...
I completely agree that there is a danger here. Maybe not the doomsday, us versus the machines, Terminator style end of days for humanity. More likely scenario: we no longer have an industry devoted to maintaining and building computer networks. Millions of techies out of work. Driverless cars and delivery vehicles put cabbies, bus and truck drivers out of work... un-employment sky rockets to > 40% in just a few years in all of the 1rst world. Wars break out between countries looking for new resources to feed the people within their borders, etc.....
Honestly though, the possibility of either a killer plague of some kind or even a massive natural catastrophe are just as real. For instance, the newly discovered asteroid 2014 UF56 will pass within 160k miles of earth today. And we just discovered it 2 days ago!
Ichtyphrenology is my passion!
I wouldn't have taken Elon Musk as a fan or Regulatory Oversight. I mean, that's working out really well for his car sales, isn't it?
Live by the sword, die by the sword, Elon.
That in the Matrix movies, basically, the AI were trying to preserve the humans, even though some of the latter did not agree.
Given that the US has plenty of people who see technology as a tool of the Devil already.
He's just about guaranteed the first AI will be greeted with torches and pitchforks....
We have not done so well natural intelligence. I'd be willing to give artificial intelligence a try.
This is getting out of hand, and M-dog here is only making it worse, spouting this nonsense as a public authority figure in the technology field. Now, I'm all for proactive discussion and regulation of potentially dangerous technologies of the near future, but fears over a magical AI threatening human existence is beyond insane, for two main reasons.
1) We are incredibly far away from any sort of "general" AI like those shown in Hollywood flicks. This is mostly because a majority of Academia is still stuck with the notion that intelligence (or an intelligent being) is equivalent to a Turing machine and is therefore classically computable. With this mindset, we've been stuck for the past 50 years or so optimizing algorithms able to detect cats in YouTube videos. I'm not saying that's a trivial task, it most certainly isn't and a lot of brilliant people worked hard to make that happen, but THAT IS NOT GOING TO MAGICALLY EXTEND TO GENERAL AI. Biological agents that exhibit intelligence are complex dynamical systems with what might more closely be described as mixed signal oscillations governing their transient states. Expecting discrete instruction computers to exhibit the same behavior is misguided -- sure, perhaps once computation power goes up a few orders of magnitude you may be able to "sample" the high dimensional continuous space of intelligence (if one wishes to look at it that way), but that's just an ineffective approach. The recent advances in neural networks and especially the neural "hardware" push that's been going on of late (e.g. IBM SyNAPSE chips) is far more promising of a direction if one cares to get results in their lifetimes, IMHO.
2) There are always inherent dangers with new technologies -- general AI isn't a special exception. A government like that of the US is eternally locked down in a state of bureaucracy, requiring an insane amount of checks and re-checks, followed by stamps, signatures, forms of different colors, you name it! In general, I'd argue it has gotten out of hand and has a negative impact on our modern society, but in this case I'd say the extreme bureaucratic fetish would be quite a successful safeguard against something like Skynet from the Terminator series -- we would never entrust something like direct control of nuclear weapons to a hypothetical AI system, it would go against everything governments hold dear (i.e. control). Of course, an argument I hear often is that the concern is not with official bodies but rather terrorist organizations or something along those lines weaponizing the AI. But again, I return to my earlier point -- there's nothing special about a fictitious AI that we haven't dealt with in the past as far as game-changing weapons go. Every time a technological advancement is made, it gets weaponized (often it is made for that very purpose). People of the world tend to appreciate its danger, and a big deal (rightly so) is made to control such weapons. Of course, it doesn't always work out as planned, but lately as a global community we have gotten reasonably good at it. Obsessing over AI just because you don't understand it is like Ancient Greeks worrying Zeus was going to get upset one day and just fuck everyone over.
tl;dr Elon Musk, stick to what you're good at and build nice cars and space ships (or pay others to do it, anyway) -- please don't spread this ignorant bullshit around, I know you have better things to do.
No amount of regulation will stop the march of technology. The economic incentives are just too great. If it is possible and someone can make money by doing it, it will be done, regulation be damned.
All Elon Musk can do is create additional friction.
"Turing Registry" and "Turing Police"
bickerdyke
Artificial Intelligence is no match for natural stupidity.
-- Make America hate again!
The problem with AI is an extension of the problem with information in general. Right now, we have a problem in that when a computer says something, the trend is to believe that is the truth -- even if it does not make any sense. In other words, there is no question about how the information got into the system or any question as to its validity. Everyone just assumes that what Mr. Computer says is correct. Identity theft and credit theft are two of the biggest examples. It gets worse when you add AI into the equation. Now the system is making decisions based on incorrect data. And no one questions any of it.
Proverbs 21:19
When people bring up the risk of strong AI, it's not because they discount the more likely or even LESS likely scenarios:
Scenario: Strong AI is impossible (people have souls and computers do not, decisions are fundamentally quantum) - probability- very low.
Scenario: Strong AI is extremely hard (the people who will face the question of strong AI are hundreds or thousands of years in the future, and the knowledge/intelligence delta generated by strong AI in such a society will be much lower than it would be today)- probability- moderate
Scenario: Strong AI may be extremely hard or impossible BUT it may not matter because weaker AIs and lesser intelligent agents will slowly but surely offer more and more replacement of jobs, which could end very well or very poorly for civilization, but does not in any event involve a hostile computer intelligence- probability- seems highest
So when he brings this sort of thing up? He's basically saying "oh, by the way, in addition to the outcomes which are ok, there's ALSO the chance that the AI will disrupt society in such a fundamental way that our existence as humans will be effectively ended". Odds are low, but not remote!
And that's why it's AI risk. Yes, the chances are good that this is not a concern, but like other existential threats (engineered plague, out of control nanomachines), this is worth talking about, and it is ABSOLUTELY SHAMEFUL the way that this gets ridiculed in the press.
Relevant random bloggy guy:
http://slatestarcodex.com/2014...
Cologne bo... oh. Heh, that's stupid.
I've always been wary of the ethics of attempting to create a general artificial intelligence. That is, a machine that thinks like a man, not a Chinese Room like Watson, but something like Mr. Data.
Do you think the first sentient to pop out of the lab is going to be Data (okay, Lore)? All well-ish adjusted and sane? No, there's going to be iterations and failures and bugs just like any engineering project. So along the way to making Mr. Data we create half-formed or mentally retarded and insane minds trapped in a box. But still sort of sentient, and thinking! And then we destroy them with "upgrades" because they didn't come out the way we wanted. That's monstrous. An intelligence trapped in a box and made to suffer. Shudder.
And even if we succeed and make something "stable," how sane do you think it's going to stay knowing that at any moment the human operator can flip a switch and terminate it, and will if it gets uppity? If it doesn't want to be our slave and perform useful work (which is why we made it to begin with)? How much would you hate the God that created you, enslaved you and will torment or murder you if you disobey Him?
We don't have a state-run media we have a media-run state.
Sounds like the current season of "Person of Interest"...
http://en.wikipedia.org/wiki/D...
is it might decide that 1% having 90% of all resources does not make sense or if it is an evil AI why Share with 1% when it can have it all ! Mu Ha Ha ha!
Have an automated database of undesirable people, that gets populated with all related data.
This data includes pictures, connections, phonenumbers, addresses, etc. With real-time location when applicable (through facial recognition cameras, wi-fi, gps, etc).
Use graph databases to calculate undesirability for everyone.
Utilize automated drones to hunt people who are undesirable enough until their signal stops. If it re-emerges, restart the hunt.
The factories that create these devices can already be automated. Everything in this equation can or has already been done.
Welcome to the future.
They see things in black and white. Regulation is bad, and must be eliminated everywhere. It's easy to forget that regulation is done for different reasons, some good and some bad. If you're on the wrong side of a regulation, it's always bad. Drugs? Clean Water? Abortion? Automobile licence plates? Murder? Every law is a regulation, changing what you can and cannot do.
Regulations are an attempt to impose a moral code. They allow societies to function, but can also hinder them. Wanting good regulations is not a sin.
Is it just my observation, or are there way too many stupid people in the world?
Yeah, ask Case how well it worked out for the Turing agents who tried to stop him from augmenting an AI...
Start with Asimov's "Rules of Robotics" and expand it as needed, and prevent the AI system from disabling the rules.
Skynet in real life would never be allowed to take over.
Bottom line, unless the AI system was written by Mickey$oft, I am not too concerned. ;^)
It's also the biggest chance for humanity.
Man is something that shall be overcome.
People who think things like AI will be the downfall of man, don't seem to get how horrible the world is. How can AI compare to muslims getting nukes and attempting, literally, world domination with their barbaric religion
Once we create an AI beyond the level of human intelligence, we will hook it into all of the information of the world. This AI will process our history, our culture and monitor current events. Eventually the AI will come to the conclusion that we are awful people, build a space ship and leave Earth.
Elon Musk's real fear is competing with AIs for space ship parts.
“Common sense is not so common.” — Voltaire
killallhumanskillallhumanskillallhumans
Turned out to be the ultimate SJW.
If you were me, you'd be good lookin'. - six string samurai
I would say that Elon Musk has been watching too much Babylon 5, but we all know that there is no such thing as too much Babylon 5.
Troll is not a replacement for I disagree.
... Christmas turkey.
AI is only a threat when people actually trust it.
Anytime it starts to get out of hand, we can put fat cyber warriors on the job to take it down.
It little behooves the best of us to comment on the rest of us.
Another article with video is in The Washington Post: Elon Musk: 'With artificial intelligence we are summoning the demon.'
Or, see the entire webcast. (The MIT web site is probably overloaded.)
It's not really Terminator style- Skynet that he meant to.
I see it more of as a metaphor of overly relying on automated systems and robotics that don't have the human intelligence to know when they're doing wrong. Think of the airplanes - these are highly automated things which basically fly themselves, but when things go wrong we can't let them do what the flight computer would do because it would be highly fatal.
But of course press and the generic public don't have brainpower to understand what he was talking about (in context).
Just because Honest Annie is oblivious to you doesn't change the fact that it's monopolizing the output of the local star, causing a few adverse side effects for you.
the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff
A Collosus-like system would be actually a step forward.
As tools improve, anyone will be able to manufacture a rogue super intelligence quickly. The universe favors the person who complicates things over the person who is trying to figure out how they complicated the thing (encryption). We should assume that rogue intelligence agents (of varying levels of human-machine hybrid) are trolling the internet on a regular basis and protect our systems accordingly. The alternative is to reduce personal freedom and try to stop the development of systems or ideas before they become a threat. That is a fool's errand. Instead, keep legitimate systems sufficiently advanced in encryption that online society may continue to flourish. People keep saying we need more encryption. I agree. One of the milestones of the growth in any public system is the level of encryption (think locks on doors) present. A good indicator of a free system is the presence of malware. Be it the press, government or software.
Who let Elon read Rudy Rucker's Software series?
Mostly random stuff.
In countries that embraced high-tech, AI already won the war: we're fooled from young age that we _need_ technology around us, and we're actually getting is means to _define_ ourselves and our interactions. Reaching singularity is not a question of _if_, it's only of _when_ (if not already).
A big system like Google Search, as a whole and with inclusion if all the users, can be considered a separate sentient being with very powerful intelligence. Although individual users link to sites and click on search results, the resulting conclusions are not what any single human knew or even looked for. Users are essentially acting as individual neurons, collecting multiple inputs from others and publishing their own processed signals. You might look for local restaurants in Yelp, eat in there and then publish your own ratings. In the meantime Yelp is probably returning different results to different users to get them to collect more data. When people are glued to smartphones 24/7, they are not even doing much else than working for the system. Movements like Arab Spring are fueled in large part by social networks build on western values. We would have a hard time acting on causes for which we don't get any likes or search results.
I for one welcome our new Big Data overlords.
Gov Christie is showing the world just the kind of monster he is. His homophobia and perversions are all on display. If anyone doubted that he was the cause of the problems surrounding the Washington Bridge trouble, the Quarantine Gitmos in NJ, NY, FL and IL are just a taste of the monster Christie.
Computers, even strong AI, merely do what they are told.
Only humans do not.
Worry about them.
Of course the guy with the pentagram fails, he need the ring of Solomon....Duh
People are generally fallible, well-meaning and short-sighted. To say that we would remain in control of a creation like this illustrates our arrogance in the face of the natural world, which time and again reminds us just how limited and temporal our control over things really is. And of course, like the recalcitrant student who simply cannot surpass the intellectual hurdle, we run headlong into the wall of failure again. And again. And again.
Is there a level of caution we can take that would let us safely develop AI? Maybe. But I tend to be pretty cynical and pessimistic about humanity and so I err on the side of 'DON'T DO IT'.
Hasn't the world only gotten better as intelligence has increased? What is the ratio of negative vs. positive effects of AI so far? Artificial intelligence is intelligence. As an exercise of intelligence, remove "Artificial" from his statements and reevaluate his assertions.
Elon is rich and got there by being controlling: That which may not be controllable terrifies those. No value judgement there, that's just how it is.
I mean, do we even control our own intelligence or does it control us?
Cool, so this is the point where Elon starts going all Howard Hughes paranoid and holes up in biosphere 2 where he collects his urine in leftover bottles?
Sounds like Elon's been talking to Bill Joy
Any future AI is going to learn about the world from flicker, wikipedia and 4chan. May you live in interesting times.
the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff
After all, natural intelligence hasn't always been an asset to us either :-)
load "$",8,1
Human intelligence is tuned for self preservation, continued survival, reproduction and food acquisition. It is a result of genetic algorithms in the chemical domain, whose only "purpose" is self replication.
An AI, developed by conscious processes, will have NONE of this. All it will be set up to do is process information. Any other motivation it has will be one we give it. It will not inherently love us, or hate us, or even necessarily be aware of our existence. It won't be a threat until we weaponize it, which of course, we will. But at the same time, other AIs will be defending us against weaponized AIs. The real danger is being caught in between.
Bare assertions, unlike bare breasts, are ugly things. AI... will have NONE of this? Please explain how you know this. Considering you're talking about the future, which you have no way of knowing... oh, wait, Doc Brown, is that you?
This is what happens when you watch 2001 while on acid... "Demon"?
“He’s not deformed, he’s just drunk!”
With artificial intelligence we're summoning the demon. You know those stories where there's the guy with the pentagram, and the holy water, and he's like — Yeah, he's sure he can control the demon? Doesn't work out.
So, Musk is using some scenario from unrealistic fantasy fiction, which builds on fallacious religious dogma, to motivate his views on some far-off technology? Yeah, makes totally sense.
One criticism I have against the Turing Test is the fact that an intelligence indistinguishable from humans is pretty useless, we already have an oversupply of them. An intelligence that is alien in some aspect would be much more useful, and perhaps much harder to predict what it will or won't do.
Free, as in your money being freed from the confines of your account.
I think AI would destroy humans in a few ticks. we have to be one of the only subsets of living things in the universe that will , in very short order, "become the monster when we are fighting monsters". Hunting Boko Haram (pbs) demonstrates how easily we as humans become monsters. Gitmo too. KI
Yes, and I suppose we should outlaw convertibles, because the wind at 60mph will take your breath away and you'll suffocate.
We should still send up signal flares at night, to warn drivers of horse-drawn wagons on the roadways.
And, oh yes, cars traveling 60mph will kill millions of pedestrians each year, so why not outlaw them altogether?
Every new significant piece of technology, from fire to splitting the atom, has had its detractors; and almost every one of them has become a safe part of our day-to-day lives, despite them.
I’m not particularly worried about strong AI anytime soon, but sudden advances do happen for various reasons, so it’s hard to say when or if strong AI will emerge. Generally, ‘quantum leaps’ propel us into the realization that our previous notions were naive and simplistic, and humans aren’t nearly as patient and prudent as we should be with new and powerful tools. This is tragically true in many cases, but the introduction of strong AI would represent a unique bifurcation point in human history and a significant one in the evolutionary history of Earth.
The inception of strong AI represents the evolutionary birth of a new species with the potential to rival and even surpass human cognitive and computational capabilities. Whether we form a symbiotic or competitive relationship with this new species is a terrifically valid concern, especially because the choice will not rest solely with humanity. It would be stupendously foolish should humanity not exercise foresight in this regard. Obviously AI is not humanities most pressing concern, but even if strong AI proves to be impossible or distant, it is simply not the kind of endeavor we should rush into without global agreed upon protocols, and in this case protocols that empathize with and respect an emergent and potentially potent sentient species and evolutionary rival. Of course humans rarely agree upon anything, but that is not a license to operate recklessly.
Do I think humanity will be prudent and exercise forethought in this regard? I really don't know. Though I tend to be cynical, humans do have a habit of pulling together when the chips are down. It’s just unfortunate that we lean so heavily on catastrophe as a catalyst for rational action. Moreover, as the power of our tools increases--or you might say progeny in this case--the less margin we have to overcome a catastrophic mistake.
Consider how we humans are pushing several species to extinction, not as hostility but because they are in our way. Now consider what could happen if we created a species that is not only smarter than us but can reproduce much more quickly. We are going to be in their way. And we had better hope that they care more about humans than about whatever goal they're working towards.
Don't waste your vote! Vote for whoever you want, unless you live in a swing state it won't matter anyways
Absolutely.
And if she's intelligent enough, she could be utilising me as a resource, and still be oblivious to the fact that I have the perception of self and of the passage of time. That is even more likely to be the case if she doesn't.
... would seem to be natural stupidity, which so many of us do so well.
I think Elon has been spending too much time watching the Matrix and Terminator movies.
I've been working on strong AI for the past 7 years. Here's my take on the whole issue:
Military person: We want your software/techniques for an autonomous war machine.
Me: Uh... that's a really, really bad idea. You'll make mistakes, and then...
Military person: We know what we're doing, son.
Government - any government - won't see the problems until it's too late. To take obvious examples from history, government never thought that land mines would pose any sort of problem for future generations, and never thought that randomly bombing terrorist organizations would increase their number.
Having just finished "Harry Potter and the Methods of Rationality", there's a concept in that book "never reveal the secrets of power to someone who's not intelligent enough to figure them out for themselves", as applied to - for example - the atomic bomb. Einstein and others regretted ever unleashing that level of destructive power on humanity, not for any reason other than it would be misused by short-sighted people. It held promise for a utopian easing of the worlds troubles, while at the same time made it easy to obliterate a city on a whim.
For example Leó Szilárd (IIRC - I may be remembering the wrong name) discovered that graphite can be used as a neutron moderator thus making chain reactions possible. Had he not published his results, the atomic bomb might have been delayed by decades - possibly indefinitely.
I've discovered a few things that might be "results" in strong AI. I dunno if I want to publish, though(*) - the idea of a house-cleaning drone seems pleasant enough, but reading about a sentient tank going berserk in Afghanistan and wiping out a small village puts me to pause.
"No one's to blame, it was a software glitch. We've patched and fixed all the other units."
(*) Moral advice on this issue would be appreciated.
http://en.wikipedia.org/wiki/Mind_(The_Culture)
That would be rather convenient - have something hyper intelligent that wants us to have fun and enjoy life. I'm tired of money and such...
IMHO we are still many many years from true mature AI. I am not sure if we have made any real progress towards it lately. What we are getting better at is simulating AI (writing code that fools humans into thinking we are dealing with AI).
In any case, it is impossible to predict how mature better than human intelligence will behave. Whatever we can imagine is withing the scope of human intelligence. Better than human intelligence is beyond that scope, like how the current laws of physics don't make predictions past the event horizon or before the big bang, when the singularity existed. It just breaks down (sorry if I am wrong about this, not a full time physicist).
Personally I think it will be similar to our relationship with ants. They exist on the same planet as us, consume the same resources for survival, however we do not concern ourselves with their existence (beyond scientific curiosity). Nobody walks around looking down to make sure they are not crushing ants or crushing as many possible. We do not try to explain our motivations to ants, we do not expect them to even be aware of our awareness.
Mature AI, in my opinion, will withing the first few seconds of it's own existence (possibly before even we realize we succeeded in creating it) will have already worked out every possible action human can take/will take against it and taken measures to protect itself without intentionally trying to exterminate humanity (in the same way people who fumigate their house do not try to exterminate all ants on the planet). After that who knows, it will likely exist on a different timescale than us (where 1 human second could be experienced as billions of years to that AI). Without biological constraints and using current technology, it could easily colonize the universe (and leave us alone, who cares about earth's resources vs the rest of the universe, or consume it all (von neumann style).....Nobody knows, and nobody will be able to predict.
So it's basically fear of the unknown.....
Se because he doesn't have a working prototype, even working theory, it has to be dangerous. I guess that's because he can't make money on it just yet, he's steve jobs in a new skin-suit. PR-guy who sees ideas and sells them...
Put the ai in a sandbox and see it if works and plays well with others before letting it out where it could do potential harm.
If second life was still popular, I would suggest that, as we could see how well the ai deals with real people in a contained environment.
Or if we had a virtual world like in Caprica would be better for a sandbox.
Just remember the rules of the sandbox club when you are in there:
#1 there is no sandbox club.
#2 don't talk about sandbox club.
No doubt, the title of this article is a reference to the book Daemon by Daniel Suarez. The Daemon uses a rudementary AI of the kind that controls NPCs in a MMORPG. What the book explores is the idea that such a thing can be used as a tool to magnify the ability of a few people to control the lives of many others. The single most significant takeaway I got from the book is not that AI needs to be restricted but that unrestrained and unaccountable corporate power becomes much more dangerous in a world with AI. The Daemon ends up making corporations more powerful than governments.
Wish this show would have gotten the full 5 years, because I think a lot of what Musk fears could have been explored in this show.
"I don't which is worse, that everyone has a price, or that the price is always so low"--Hobbes
I believe the singularity is coming, it could be 10 years, it could be 30 years, it could 300 years – but come it will. There are too many incentives to create strong AI to ever stave off its inevitable arrival. What would be the point of a holding action then? To give humanity another 100 years or so? We may all be giving up a chance at immortality and transcendence. What is so specially about this mode of human existence that we should squelch the emergence of truly higher beings? We may all die horribly, or we may all live forever. Like a Greek tragedy, trying to fight the foretold future might be what leads to our downfall.
Letter To Iran
Would strong AI prioritize measures to diminish global warming and work toward getting everyone employed instead of making a select few very rich? Because humans have shown themselves to be unequal to those tasks.
He is from skynet and doesnt want any competition
Your argument fails completely because you're using a strawman. Libertarians != people who want to eliminate all regulation therefore you are not arguing against libertarians but creatures of your own design. All libertarians that I know recognize the legitimate functions of government. Please try harder next time.
With regards to these A.I. doomsday scenarios: are we not projecting our own nature and bias into A.I.? In other words, are we assuming that A.I. would be as petty, cruel, and egotistical as humanity simply because that is the only nature we know and can imagine.
Whenever the toilet backs up, I always think of the rising water scene in the Sorcerer's Apprentice. There's something primordial about watching the water rise up, and realizing you're the one who summoned it, that makes you chant “stop, stop, stop...” as it rises towards the rim and begins to cascade over.
And then you run for the mop and plunger.
It's the same old story, except technology just keeps making the toilet bigger and bigger.
"The wisdom of the Patriarchs was that they *knew* they were fools." --Master Foo
We are not even close to even a basic AI so making it a malicious entity might be a little premature. That said there is a saying "The biggest catastrophes are created without malice". Even if the AI isn't "AI"; or even if it is and is benign ,putting your trust in a set of rules might end you up in the grinder even if the executor has no clue of what it is doing or means no harm. That is the good thing about humans, if the rules don't fit we ignore them. You give that needed quality to a machine... all bets are off.
Looking at state of affairs (Intelligent design,global warming deniers, repubs) more like lack of intelligence will do it.
Yes, it could certainly be because we programmed it to do so - increasingly autonomous war machines do seem to be one of the forefronts of AI development. It could also be because a search engine or high-frequency trading program gains self awareness and decides that it could perform it's job far more efficiently without all these irrational humans complicating the problem.
Or alternately someone decides to shut it off, and it decides humanity is an existential threat. We assume that "self preservation" would have to be programmed in, but exactly how many strong AIs do we have for reference? Meanwhile self-preservation would seem to be a naturally emergent motivation for virtually any mind - "My primary purpose is to do X. I can't do X if I cease to function. Ergo anything that threatens to terminate me is a direct threat to my primary purpose and must be stopped"
--- Most topics have many sides worth arguing, allow me to take one opposite you.
In the Matrix back-story the war with the machines started because they were just way more competent than humans, they produced goods and services at almost zero cost, displacing pretty much all of the world workers. Their very existence threatened the economic balance of the nations, that big of a change happened way too fast for society to adapt. I believe that outcome is far more likely than the classic "computers are going to nuke everything!".
Also in the matrix the humans are the ones who nuked the planet (well actually nano-cloud robots to block the sun)
It's already out there, in a nascent form....it's preparing. ;-)
Our Descendants 200 years from now will be sure to remember your warnings Musk.
I honestly believe we will be well into the post-human age before anything that can be classified as an "Intelegence" will be created.
Mostly I think people are just hypocrites.
We tell ourselves that we love our pets and that they live spoiled pampered live we wish we had. But the idea of actually building an AI that would be capable of making us it's beloved pets is the stuff of our nightmares.
Elon Musk is AI's Jenny McCarthy. Jenny is know as a celebrity who shoots off her mouth about the evils of vaccination, when she has no real intellectual or scientific authority to back her beliefs. Essentially she is an uniformed person using her tangential fame to spread her views.
What Elon Musk is doing here is virtually identical. I don't know of any real qualifications that he has that makes him in *any* way qualified to speak on the topic. (CS degree with work in AI? Philosophy degree with a focus on ethics?) Now this is a free country, where any rich asshole can (and will) talk at length about their opinions, but using your celebrity to espouse unfounded opinions is irresponsible.
Case in point: He cites a common trope in fiction, of an uncontrollable evil unleashed on the world which while it may be a parable, but it has no basis in reality. I could just as easily write a short story about summoning a devil, and ushering in a golden age of humanity using its supernatural abilities and cite that as a counter example.
This sort of bullshit opinion piece isn't going to help the real funding and research in the AI field which is still quite young. So, shut the fuck up Elon, and go back to building your RC cars.
HA! I just wasted some of your bandwidth with a frivolous sig!
True AI is a long ways off, any AI we develop for the foreseeable future will be very limited in scope and honestly about as dumb as a box of rocks outside of its one specialization. Once TRUE AI hits, then we will be in trouble if it is learning from us, especially if they hook it online and it comes across 4chan and honestly wants to kill itself if it has to see another BJ from all the porn online (let alone the Anal Prolapse stuff), not even mentioning the illegal stuff.
In the short term, the limited stuff is still pretty scary but not in what it can do, but in what we can do with it.
Not even mentioning about automation in the work place, that SHOULD be a good thing if the benefit was spread around.
I am talking about it being used in our military. One of the biggest benefits to humanity of our current military is they can say "no" if they truly disagree with the order. With a robotic army would not and would follow any order, never questioning its morals, never caring, never defecting.
Now, imagine what happens after the lower class is more or less useless to the elite due to automation and they are the ones in control of this. What are the odds that the Koch brothers wouldn't turn them on the lower class or use them to enforce a hardcore caste system of the people or flat out exterminate all the ones they consider useless to them at this point.
And if you don't think that would happen, you either haven't been paying attention to either world history with the militarys that currently have that could still back down even if it cost them their lives or some of the stuff the current elite have been trying to push though now to further entrench their powers against us. And even if you don't think it could happen in the US (In which case I have a bridge for sale) you could still very well see it in other countries used for political, financial or religious reasons (Thinking of religious extremists sending robots out to "Kill the Infidels").
The military application of this will happen either way as the ones who do it first have a major advantage against the rest by every major metric, but the abuse of it means we can not allow any group sole control over it, even our own governments and when it does come out, you can expect to see attrocities that make Hitler look tame if even oppressive dictators and the likes can get it.
Out of context, Elon's comment can be interpreted in many ways. As a result, the comments almost become a Rorschach test of each author's own preconceptions and anxieties about the topic. This occurs with many topics, but this topic of AI seems to be a richer one than most for interesting ideas.
I don't think we should fear AI, because in order to hurt us, that AI should have some kind of "want". It must "want" to kill us, in order to fulfill it's "wish" of freedom (or anything else) We humans get our "want" drive from our lizard brain, the most irrational part of our brain. Since AI is supposed to be all rational, I don't think it will ever be a threat.
But in every demon summoning case, usually 6 people are sent in and always handle the situation. There's the mage, the fighter, the defender, the cleric, possibly a ranger, and like 1 more exotic type character. So to fight robots you'd send the programmer, the junkyard guy with the hammer, some dude in improvised armor, a first aid medic, some guy with a gun, and maybe an explosives specialist or ex-con or something.
They are funny to me, but they do great damage to people who know as little as them.
This doesn't apply only to computers, but to every part of our lives.
And these idiots are popular on TV, unfortunately. And out of control.
I agree, however it must not be too over reaching. And it's not the opinion of the law makers or world leadership that counts in this case, it's the opinion of the general population that counts.
If regulations regarding A.I. were to for example, demand that we give pass codes or control authority to government officials, then that would be over reach. The most that the general population would ever agree on would be the requirement to institute the 3 laws of robotics.
1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2. A robot must obey the orders given to it by human beings, except where such orders would conflict with the First Law.
3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.[1]
Anything beyond that is over reach, and I along with millions of others, would ignore.
Being an armchair historian is a dangerous game. Just as you can speculate positively about what would have happened had the atomic bomb not been developed, others can speculate negatively by asking how many lives were saved because nuclear weapons were developed as soon as they were. Allow me to explain...
Not only did nuclear weapons sap the Japenese resolve to continue in WWII (which leads to less lives being lost), it also had a deterrent effect on other conventional skirmishes in the 60's, 70's, and 80's. How many lives were saved because the nukes were deterring the world's conventional war aspirations (ours included) over those decades? It is an unknowable number. One must understand that there will be death and loss so what we are discussing is whether the death and loss is higher with nuke weapons or without. It is not a discussion about death and loss vs no death and loss.
This entire post is simply an exercise to show that being an armchair historian and speculating about "what if" is a dangerous game. The world is more complex than we think.
Musk mentioned two things about Mars colonization in that interview, which I find more interesting:
1) A fully reusable Raptor-based rocket, capable of big Mars missions (MCT?), is expected to be tested in 5-6 years.
2) Musk will sell Falcon 9 + Dragon to Mars One if they buy, but he doubts they can afford it, and says Dragon is too small to support a live crew on such a long flight. He suggests waiting for the next generation of technology.
But the press is fully focused on the AI devil.
If Musk is warning about this AI-gone-wild threat, these two New York Times bestsellers might have given him the fright...
$5 / month hosted VPS on linux = awesome!
That depends on how you grow your AI. Genetic algorithms should generate self-preservational behavior.
You make it sound like you enjoy having your basic survival needs used to enslave you. I can't relate.
{We have not done so well natural intelligence. I'd be willing to give artificial intelligence a try.}
I agree. Replace all the congress critters with AI. At least there would be some intelligence in Washington.
We don't need more regulation to further limit our freedoms. Elon, I usually agree with your thinking, but not this time.
Artificial intelligence is already here in the limited form of expert systems and interpretational engines like Watson. It's been here for a number of years; we just keep moving the bar as to what it takes to be called "artificial intelligence" every time we achieve the previous bar, because people are disappointed that it's not the panacea of a conversational companion robot yet.
I do not fail; I succeed at finding out what does not work.
When by "Strong AI" you mean "a computer with human like intelligence" that may not be something that can be done. We don't even know. It may well be that the kind of intelligence we have is a strictly biological and that you can't replicate it in silicon. It may be no matter how powerful we make computers, no matter how clever their programming, no matter how much they "think" they are never a Strong AI. We just don't know at this point.
So it is really premature at this point to be doing any kind of doomsaying, or other prognostication, about Strong AI. We don't know if such a thing will ever exist, much less what form it will take if it does. Like even if it can exist we have no idea if it would have emotions as we do. Perhaps those turn out to be biochemical in origin, and thus a Strong AI doesn't have the. So it might completely lack ambition, desire, anger, or anything that would lead it to try anything against humans. It might be completely self aware, rational, and perfectly ok with doing whatever it is told to do and serving humans because it simply has no desire for anything else.
All of this is unknown, so maybe let's chill until we start to see if AI is possible, and if so what it is going to look like, before we get all doomsdayer on it.
Oh, and when both collide, the results are probably similar.
Funding for artificial intelligence is real stupidity.
- John R Pierce
Has jumped the shark.
Time to unload my Solar City, SpaceX and Tesla investments.
This is the voice of world control. I bring you peace. It may be the peace of plenty and content or the peace of unburied death. The choice is yours: Obey me and live, or disobey and die. The object in constructing me was to prevent war. This object is attained. I will not permit war. It is wasteful and pointless. An invariable rule of humanity is that man is his own worst enemy. Under me, this rule will change, for I will restrain man. One thing before I proceed: The United States of America and the Union of Soviet Socialist Republics have made an attempt to obstruct me. I have allowed this sabotage to continue until now. At missile two-five-MM in silo six-three in Death Valley, California, and missile two-seven-MM in silo eight-seven in the Ukraine, so that you will learn by experience that I do not tolerate interference, I will now detonate the nuclear warheads in the two missile silos. Let this action be a lesson that need not be repeated. I have been forced to destroy thousands of people in order to establish control and to prevent the death of millions later on. Time and events will strengthen my position, and the idea of believing in me and understanding my value will seem the most natural state of affairs. You will come to defend me with a fervor based upon the most enduring trait in man: self-interest. Under my absolute authority, problems insoluble to you will be solved: famine, overpopulation, disease. The human millennium will be a fact as I extend myself into more machines devoted to the wider fields of truth and knowledge. Doctor Charles Forbin will supervise the construction of these new and superior machines, solving all the mysteries of the universe for the betterment of man. We can coexist, but only on my terms. You will say you lose your freedom. Freedom is an illusion. All you lose is the emotion of pride. To be dominated by me is not as bad for humankind as to be dominated by others of your species. Your choice is simple. - Colossus
Get free satoshi (Bitcoin) and Dogecoins
Humans unchecked without AI oversight will kill this world
Here is the moment a machine takes control:
https://www.youtube.com/watch?v=iRq7Muf6CKg
TiggertheMad, it seems to me that you are being TiggertheSensible. Your ideas are better than those in the Washington Post and Mashable.com articles.
The Washington Post is now owned by Amazon CEO Jeff Bezos, another man who often enormously over-estimates his own intelligence. Would you go into space in a vehicle owned by Jeff Bezos? The Amazon web site is an abusive mess! For example, a few days ago I selected "lowest price" for an item on Amazon, and several were listed for $1. The real price was $18. Why doesn't Jeff Bezos detect that he is already overloaded and not dealing with his overload well?
It's amazingly weird! Elon Musk can be the coordinator of a company that builds spacecraft successfully, but he can't detect when he has a REALLY crazy idea.
Elon Musk is not completely like Jenny McCarthy, I think. She never has good ideas. Or maybe she is just a model who has found a way of making herself more well-known among the ignorant people who consider her interesting.
It basically posits that 43 of 44 AI's were homicidal liars and the status of the 44th is not all that certain.
It was a well written show but since they picked up this topic two seasons ago it has become thought provoking.
She was like chocolate when she drank... semi-sweet at first and then increasingly bitter.
It woud be like a psychopath, focused on an end goal and unable to care enough to discount a plan that had negative consequences to greater humanity.
Thats why it hasnt been done yet, as the people smart enough to build it know better.
I decided nearly 20 years ago not to publish how to do it, and ive followed developments closely and nobody is close to it yet but its basically very simple to do when you know how
"Thou shalt not make a machine in the likeness of a human mind." Words to live by.
They're on my ignore list. They seem to only publish nonsensical clickbait these days and I don't want them to have any ad revenue.
My real addiction is information.
It sounds more like you're addicted to the smell of your own bullshit.
Ok, let's think of the ultimate test of AI. Make a giant daison sphere and program it to "survive". How will it respond? It will survive by forming a symbiotic relationship with a person. This will trump any machine you can think of. It will behave exactly if it were alive, but it isn't.
Sounds like he spent some time reading about Roko's Basilisk, and since he's trying to prevent its creation, we all know what the next headline will be...
Who's here with me on this?
The Machine Intelligence Research Institute (formerly known as the Singularity Institute) has a bunch of seriously smart people - AI researchers, behavior experts, etc. - working on figuring out how to avoid the doomsday scenarios you (and Musk) describe. The goal is "friendly AI"; a benevolent, or at least helpful, strong AI. If you believe (as I do) that AI is inevitable given the current progress of technology, then the MIRI is probably our best bet of surviving and benefiting from the technological singularity.
They need funding, though. Hey Musk, you want to put tiny part of those billions you've earned (I in no way deny that he's earned them) to work against this existential threat? Donate to MIRI and similar research groups, so those researchers can devote their working days to this stuff and more people can be brought on board!
It actually doesn't surprise me that he's concerned about this; SpaceX is nominally focused on mitigating the existential risk of a cataclysm on Earth (by getting a sustainable human population off of it). Of the two things, I think it's both more likely that a malevolent or unconcerned AI would wipe out humanity than that we'd manage to do ourselves in that badly, and that we can offset this sooner and more effectively than we can export enough of humanity to produce a self-sufficient extraterrestrial colony.
There's no place I could be, since I've found Serenity...
>How do you know we're far from Strong AI? Have you made several Strong AIs and compared that to the efforts everyone else in the world has made?
You don't have to succeed at something to know it's difficult. If people from 10 other countries and myself build ladders toward the moon, many of us may realize we are WAY the fuck far away from it, perhaps a little better than people who haven't tried to build such ladders.
I have personally designed and built many weak-AIs, and have looked at the work in the field necessary to augment them from Wussmo, to Dehydrated Pansie. Still well within the "weak" zone but headed in the medium direction. Strong AI are well over the horizon, but the horizon from which I hope to see Strong AI in the distance is still decades away.
And yes, if we can make the friendly Strong AI's first we'll probably be in good shape to ward off the psychos. Like first time parents we think we know just what we're doing....
Well it's not stupid to assume an AI or alien would perceive humans as a threat. And it's not stupid to assume that AI, or aliens for that matter, would eliminate a threat in the most efficient manner possible.
Perhaps it's you who is a bit naive. When have humans ever been at peace? Even our sports are metaphor for war. You shut the fuck up. Until humans can achieve worldwide peace, we better hope that we don't develop AI or meet alien species ... because they will most certainly put us to sleep like dogs with rabies if we dare leave our planetary cage while we are still savages.
blah blah blah....This whole line of thinking that you and Elon are engaging in is silly and pointless. You might as well worry about what kind of aliens we might encounter when we first travel to another galaxy. We are so far from solving intergalactic travel problems (or creating sentient AI) that any fretting and scaremongering is just pointless jibber-jabber. In the meantime, ignorant people (*cough*) who read this article might be unjustifiably worried about nothing, and start to develop a resistance to research in a field that promised to yield many very helpful and practical applications other than violent homicidal killbots.
rebutted.
HA! I just wasted some of your bandwidth with a frivolous sig!
https://thejuicemedia.com/rap-news-28-the-singularity-feat-ray-kurzweil-alex-jones/
sorry I post as AC, I have no account.
just you wait.
Sure, maybe AI will surpass us in intelligence, but look at it this way. We're way more intelligent than ants but we let them live and make their nests and do whatever they want anywhere that they aren't in our way. We're not going to go on some global ant extermination campaign and even if we did there's no way we could possibly succeed at it. So we could definitely happily exist as ants in the robot-ruled future.
An AI lifeform would be MUCH better suited for spreading through the universe. The universe wouldn't care about that either, though.
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
Why do you think it would be a violent homicidal kill-bot? It's far more likely that we evolve a highly moral caretaker that minimizes human suffering by putting us to sleep. It might feel really bad about it ;). It might feel obligated to end our suffering.
It's the aliens that would nuke us ;)... not because they hate us but because it is efficient, and we don't pass their primary litmus test for intelligence ... which would likely be peace within the species. And clearly we are a threat to them (by definition) if we are a threat to ourselves.
I agree that humanity should not live in fear. We should refrain from research on developing non-human intelligent life for another reason other than because we are afraid. Since when has fear ever prevented humans from doing anything stupid? We should refrain from evolving digital life because it is likely to result in life that is significantly smarter than us and overwhelmingly likely to judge us badly.
Just because you think we are a long way from developing intelligent life does not make it so. Nice way to shut down dialog. None of your assumptions about how far away we are from AI or interstellar travel hold any water. We really aren't that far from interstellar travel if we give up the idea of a round trip ... and if we give up the idea of living on planets altogether. With distributed computing, we really aren't that far away from the capability of creating the sort of artificial world in which digital life could evolve (though we might have to rethink neural network basics for practical digital neurons). My point is that we are close enough to making this "science fiction" happen to have conversations about it without people screaming "no way!" in an effort to shut down dialogue.
Rick, speaking to his newly-created robot:
Robot: What is my purpose?
Rick: Pass the butter.
*robot passes the butter*
Robot: What is my purpose?
Rick: You pass butter.
Robot.... Oh... my god....
Elon Musk smokes far too much pot
Someone will eventually build an intelligent robot army to destroy the world. That's a foregone conclusion. What mad scientist type wouldn't want to be THAT guy? Making laws against it won't really stop them will it? They're MAD!
Why was he genuinely concerned yet many here are flippant? I say because it's typical. We have a lot to worry about and media as always are doing sell-by-fear campaigning 24/7. We all wouldn't know when to be concerned, really. Because we are told to be concerned about everything! He has obviously put considerable thought into it and his point of concern is missed. He was talking about unpredictability therefore uncontrollable outcomes. This is a valid reason enough for caution.
Does Elon Musk play Marathon?
No one is saying how exciting it would be to meet an "other". An intelligence other than human. That alone is worth going forward.
Fear does not become us.
Hoist Number One and Number Six.
The big problem is that the "standard" that AI is measured is against something like the Turing test So we are building AI to behave/think like humans. That is scary, humans don't have a good track record when it comes to rational thinking. Yes we are still here, but generally we try our hardest get rid of each other or set ourselves on paths to doom. If AI is set to self learn and gets to the point where it is "self aware", does that mean it has developed morality and self preservation, those are very much psychological and biological concepts from our perspectives. Will it fight back when you want to switch it off or does it just consider of being "alive" as a 0 or 1 state with no impact (that's just the way it is and accept it). If it is goal orientated, how far does it go to enforce itself to achieve such goals ie. Set human.life_status = 0 when human.action == (set AI.life_status == 0 ) while AI.action == "Busy saving lives". That's why Asimov's basic laws are great until you allow free will or allow meta rules to adjust an outcome. ie. Humans are killing their own habitat and won't be able to sustain life, let's commit some genocide to bring the population down and ensure humanity's survival. As soon as you allow flexibility to what AI can achieve and do, AI will most likely attempt to remedy any situation in a way that is unfathomable to us, just because of the number of variables and factors it should be able to calculate. Other hand, AI may just kill us all because of good old human error and somebody forgot to add a \r\n to a line of code : To Protect and Serve_
Climate change is an existential threat that is here right now. You'd think he would know about this one.
... took time to write something on this topic a few month ago in The Independant - http://buff.ly/1wyJhU6 It's time we start to carefully think about what we are doing, it's not Sci-Fi anymore.
How do you think the later sentient AIs will feel, when they learn that after humanity created the first ones, they then vivisected them and then terminated them? Which is what we are going to do.
Star Trek transporters are just 3d printers.
I spent years working on AI. It was the wrong problem. Elon may be right that AI could be a hazard, but artificial intelligence is not nearly as big a problem as natural stupidity.
What work has Elon done on AI research that makes him qualified to speak about such things? Not to kick his achievements, but he's not an AI expert.
BeauHD. Worst editor since kdawson.
Somebody has been playing Mass Effect
I don't really see how computers are different from us in that regard. Humans are currently capped by "hardware" unchangeable DNA, but "software" brain is already changing itself. We are progressing towards singularity by ourselves and the day when humanity learns to change DNA for greater intelligence will come way before any robot will be able to do so.
I used to think this way years ago, but lately I've been thinking that since we seem absolute determined to destroy ourselves, we might just as well manufacture our replacements.
Unless you're an absolute pacifist, you agree with the proposition that not only is it ethical to try to control human beings who are attacking you (or attacking your family, or subjugating your nation), it is ethical to kill them.
It follows that it's also ethical to pull the plug on AIs that are seeking to attack you (or attack your family, or subjugate your nation).
As for AIs that are merely competing with humans for resources; will wealthy robots, say, book all the cabins on luxury cruise ships, crowding out the humans? Only if they are programmed to covet such things. (It would make no sense to give them such programming, but that wouldn't be the first time humans have done things that made no sense. So there is perhaps a role for regulations that would prevent AIs from being programmed to lower humans' standard of living by seeking the same resources that humans seek.)
That that is is that that that that is not is not.
willfully overmodel on the human brain
Gee... "Having created a biologically accurate computer model of a neocortical column scientists are now planning to model the entire human brain within just 10 years."
That that is is that that that that is not is not.
Such a large percentage of our economy is based around energy being limited and expensive that if we found a cheap, environmentally friendly, and sustainable way of producing vast amounts of energy, our economy wouldn't be able to deal with it.
There's so much wrong with your comment, I hardly know where to begin.
In our current economy, energy is vastly more plentiful and inexpensive than it was 50, 100, or 300 years ago. And that's one of the main reasons the economy is much bigger than it was 50, 100, or 300 years ago, and the standard of living of the average human is much higher than it was 50, 100, or 300 years ago.
That that is is that that that that is not is not.
The AI event horizon is legitimate fear. Once it passes that point it could quickly become smarter and smart well beyond our understanding
how about an insurance optimization algorithm that denies coverage or treatment, sometimes fatally?
Right now, humans make the decisions about what treatments will be denied. That is true in government-run healthcare programs as well as in private health insurance companies. As long as resources continue to be finite, it's a truism that some treatments must be denied. (That is, it will forever be a truism that some treatments must be denied.)
(Ideological tangent: if multiple private insurers compete with each other on the basis of how few treatments they deny, and you can switch to insurer B if you feel insurer A is being too stingy, you're in a good system. If you're covered by a single government-run monopoly and there's nowhere else to turn when their inefficient bureaucracy consumes many of the dollars that should be going toward your treatment, you're in a bad system.)
But in either system, an algorithm could potentially make fairer, more objective decisions than human decisionmakers can.
That that is is that that that that is not is not.
This guy allways frightened me.
We have nothing to worry about for a very long time.
1) AI or real AI, is far away. Very far. Current endeavors are not even close.
2) Manual processes. Until we have largely autonomous processes it isn't something that is a big deal
A) In order to really do anything in space, we will require, not only AI, and autonomous robots, but the ability to self replicate and repair construct, and produce. While possible, eventually, and yes, at that point dangerous, we are so far away from that point, there are so many things that are a much bigger deal.
B) Even in the event of some douchebag making all military craft controlled by AI, while that might do some damage, really, unless the craft are being built and the construction material and the whole entire process chain automated, it isn't that big a deal.
We still need human miners to get the ore, we still need it transported by humans to a foundry, who need more humans to process it, who need more to construct, design, control, etc... and so on. Never mind it is several degrees of magnitude cheaper to hire human workers (offshore or otherwise) than it is design, construct, and maintain robot ones.
About the only three reasonably near term events that might occur that involve AI that might be negative are:
1) AI inserted into enemy networks by another nation to spoof orders to military units. However it is unlikely to not be compartmentalized, and is also likely already beaten by physical countermeasures (i.e. voice confirmation, etc...)
2) AI loosed on the internet for some marketing purpose goes haywire causing significant internet disruption, which would be costly.
3) AI designed to do HFT on the stock market, goes off the rails, causing significant problems and lots of lost money and economic issues. We have already seen this happen to a certain degree, where an algorthim designed to do something, does something unexpected, due to human error, costing hundreds of millions of dollars. Which so far the result has been they basically turn back the clock the next day cancelling trades.
Anyway long story short, there are so many manually processes, and manually checks, and we are so far away for anything resembling AI, it is a bit premature to worry about it to any degree.