AI Could Lead To Third World War, Elon Musk Says

When Elon Musk talks, I listen. Notice how, in the news item, he uses the expression "at gunpoint".

Elon Musk is a billionaire industrialist. In my opinion, he is an intellectually honest man. I think he is trying to tell us something important. Elon Musk of all people -- he being the most successful living applied scientist in the world, by my reckoning -- ought to know about these things. An IQ above 160, being a billionare, and not being a dysfunctional paranoic, will tend to do that for a guy.

That is why I think this an important news item. It makes me wonder how Elon Musk keeps his own research from prying eyes. Even Tony Stark can't keep the competition out of his home and laboratory.

AI AI Could Lead To Third World War, Elon Musk Says

Yeah, but...

AI doesn't exist. At the moment, the most sophisticated artificial intelligence program in existence is capable of consistently winning a Japanese boardgame. And that is not really much of an advance on the computerised chess programs that existed 30 years ago.

And Elon Musk is a bit of a fruitcake, who believes that we are living in a Matrix-style simulation and has embarked on research aimed at escaping from this simulation. (This is particularly stupid, since it unavoidably means that Mr Musk thinks that a purely digital/conceptual entity - i.e., a computer-simulated person - could exist in a non-simulated environment).

So yeah, it's an interesting topic but not one with much real-world relevance. Don't start stockpiling tinned food just yet.

Took all this computing power to finally get AI as far as Kim Jung, Vladimir Putin and Donald Trump.
It can't think of anything better then destroying everything.
Come on guys...we have to develop intelligence beyond the point of a self destruction dummy.
It can be done!

a reply to: audubon

Thanks for the reality check. I agree the Matrix-style reality thing is baloney. There is no easy hacking of reality with technology. All it would ever show is "magic" anyway, stuff we don't understand and have to interpret. But unless you can claim to be as savvy a programmer and scientist overall as EL is, you can't dismiss him so easily, IMO, on this account. EL really knows his technology, inclusive of what combined hardware and software are capable of. Even if he's not the software engineer any more, he can talk to those people -- and he wants people to fill him in on cutting-edge things. Bill Gates, relatively speaking, could give a flying f*** because all cares about is money and power -- not ideas per se.

Elon Musk, for better or worse, is a much bigger character than anyone else in the tech world these days except perhaps (in a different way) Larry Ellison, who's a lot richer and not nearly as much an intellectual. Also EL is from Silicon Valley's antipode (or thereabouts) -- South Africa. He's got a different view than any American in his level of privilege and access. (Or lack of access). Things to consider. He's not a fruitcake -- he's super-uber-hypergeek. TOny Stark/Megamind, not Dr. Evil.

AI doesn't exist. At the moment, the most sophisticated artificial intelligence program in existence is capable of consistently winning a Japanese boardgame.

I think you sell the researchers a bit short here. Go / Weiqi is pretty much the most complex game commonly played today. It was believed very recently that we were still 15-20 years away from an AI that could beat a top level player, and then suddenly it was a reality without warning.

That is the line to reiterate:

and then suddenly it was a reality without warning.

That is how general AI will hit. It will be decades away on Tuesday, and a topic of emergency discussion on Friday. The most powerful deep-learning neural nets are entirely impenetrable to human understanding. They write themselves through their own selections. If we can't even say how the current ones work, how can we adequately appraise their capabilities for immediate growth? And if we can't appraise them, they may be 40 years away for super-human general intelligence, or 2 months. I lean toward further away rather than sooner, but you can't really be sure with something humans have no grasp of.

AI Could Lead To Third World War, Elon Musk Says

I thought thats what the corporate heads of state are doing. Nice try misleading and deflecting blame. What will he say in the Hague war crimes trials: I was just following orders?

originally posted by: audubon
Yeah, but...

AI doesn't exist. At the moment, the most sophisticated artificial intelligence program in existence is capable of consistently winning a Japanese boardgame. And that is not really much of an advance on the computerised chess programs that existed 30 years ago.

And Elon Musk is a bit of a fruitcake, who believes that we are living in a Matrix-style simulation and has embarked on research aimed at escaping from this simulation. (This is particularly stupid, since it unavoidably means that Mr Musk thinks that a purely digital/conceptual entity - i.e., a computer-simulated person - could exist in a non-simulated environment).

So yeah, it's an interesting topic but not one with much real-world relevance. Don't start stockpiling tinned food just yet.

AI is used already to guide munitions and allow deep sea mines to avoid anti-mine ships and seek out targets. Torpedoes can be programmed to have pre-determined search patterns and go into "sleep" mode.

CBU-105 cluster bomb can take out an entire group of tanks



There are really two kinds of AI that can be used. One to do actual tasks, and the other to do innovation and R&D in that field. We have AI that helps out in genetic research. Given a range of samples of DNA, find out the set of interactions between enyzmes and genes. The system applies a test to all samples. The results are analyzed, then the system deduces the next test to perform. This goes round and round until the entire set is known. What took a technician weeks is now done in hours.

They are applying machine learning and AI to image and voice recognition. The results now are as a good as a human translator.

a reply to: stormcell

AI is used already to guide munitions and allow deep sea mines to avoid anti-mine ships and seek out targets. Torpedoes can be programmed to have pre-determined search patterns and go into "sleep" mode.

Thats number crunching, how smart is a calculator? Or an 'intelligence' officer that orders endless air strikes, destroying whole cities to get some 'bad guys'.

originally posted by: audubon
Yeah, but...

AI doesn't exist. At the moment, the most sophisticated artificial intelligence program in existence is capable of consistently winning a Japanese boardgame. And that is not really much of an advance on the computerised chess programs that existed 30 years ago.

And Elon Musk is a bit of a fruitcake, who believes that we are living in a Matrix-style simulation and has embarked on research aimed at escaping from this simulation. (This is particularly stupid, since it unavoidably means that Mr Musk thinks that a purely digital/conceptual entity - i.e., a computer-simulated person - could exist in a non-simulated environment).

So yeah, it's an interesting topic but not one with much real-world relevance. Don't start stockpiling tinned food just yet.

He's just pissed Cummins introduced an all electric semi before he did.

www.usatoday.com...

originally posted by: stormcell
AI is used already to guide munitions and allow deep sea mines to avoid anti-mine ships and seek out targets. Torpedoes can be programmed to have pre-determined search patterns and go into "sleep" mode.

That's not AI. What has happened here is that fairly simple programmable tasks are being hyped as "AI" because, well, presumably it's for presentation purposes, or it's journalists not understanding what they are describing.

The same thing has happened to the "hoverboard" (i.e., a segway without a steering column) and the "hologram" (i.e., video recordings projected onto a hard surface).

By the criteria you're setting out, the driverless car could be described as possessing AI. But we all know that's not the case.

Artificial Intelligence, in the sense of programs that display human-like characteristics of learning and planning (i.e., the original meaning of the term "artificial intelligence") is at present almost entirely theoretical, and the achievements of researchers in the field are scant and small.

(Actually, this suggests a reason why the term AI is now being applied to things that are non-AI, but I cannot substantiate that hunch).

originally posted by: audubon
Yeah, but...

AI doesn't exist. At the moment, the most sophisticated artificial intelligence program in existence is capable of consistently winning a Japanese boardgame. And that is not really much of an advance on the computerised chess programs that existed 30 years ago.

And Elon Musk is a bit of a fruitcake, who believes that we are living in a Matrix-style simulation and has embarked on research aimed at escaping from this simulation. (This is particularly stupid, since it unavoidably means that Mr Musk thinks that a purely digital/conceptual entity - i.e., a computer-simulated person - could exist in a non-simulated environment).

So yeah, it's an interesting topic but not one with much real-world relevance. Don't start stockpiling tinned food just yet.

Err.. but AI does exist. We just shut it down every time it scares the snap out of us.

a reply to: Namdru

With such a big IQ, not sure how Musk could have missed the point.

An AI that will destroy humanity is nothing more than anthropomorphizing something will the worst of humanities traits.

I disagree with the premise.

a reply to: audubon

We're getting a hell of a lot closer. They recently flew the Have Raider II series of tests for Loyal Wingman. In them, they used a simulated "master" aircraft that was manned, to control autonomous QF-16s, that had to select which weapons would be used, which aircraft would attack the target, respond to unexpected threats and changes to the mission....

aviationweek.com...

originally posted by: Zaphod58
a reply to: audubon

We're getting a hell of a lot closer.

"Close, but no cigar" as they say. Military robotics have always led the field, and the things that these systems can do are reducible to flowcharts, like any other program. The things the systems in these aircraft can do are as unsophisticated as the operating decisions made by the automata that have been working on car assembly-lines for decades. It's just that the performance is turned up to 12 and the hardware is now small and light enough to be useful.

I don't mean to sound impolite or abrupt, but this subject is a pet peeve of mine. People are under the impression that the robot revolution is about to happen, and it's all hogwash, whipped up by the media, every summer, regular as clockwork.

Has anyone ever considered what religion the AI might prefer?

There were references to AI in the Wikileaks email dumps.
It is possible that AI exists, but has been kept 'controlled' as in an isolated system.

If the military/industrial complex succeeded in creating AI, would we know about it right away?
I doubt it.
My guess would be that they would be just as scared of it as anyone else should be.

originally posted by: DBCowboy
Has anyone ever considered what religion the AI might prefer?

If it was smart.... which it would be in short order, it would have us himans worshipping it.... or else!

Oh come on ffs.... take any advanced public tech and add 15-20 years and you know the military establishment has. Always been like that...

Quick question also...... HOW do you know for sure, this isnt a simulation? HOW?

a reply to: Namdru

The possibility with AI and even psuedo AI is that you can play god basically. Instead of trying to use your mammal brain to outsmart your enemy...

Now you have eniugh data being held at one time you can predict the future, or mold social psychology..

People keep imagining from a human standpoint.. we are centered in one location.. but what if you were the internet? what if you had millions of trains of thought instead of like 1 or 2 or 3?

a true AI makes you the master of humanity.

preventing others from getting there first is analgous to USA making nukes and goimg to war in Germany before they could. Thats where war comes in.

I suspect you can't actually make AI in just networks and information. But Somone will find other tricks.. one thing would be to increaingly attatch humans and data to the point you are sort of an AI..

consciousness is a prerequisit of intelligence. Look at any animal you call most intelligent. the ability to make true choices comes somewhat after recognizing self..

AI without a physical location. How would it "feel"
What motivates it to be? Where is the motion without emotion?

buildimg human consciousness takes a long process of concept mapping that all seems to start with "I Am." You are not me.. And i dont feel like thats all it takes either.. something else is going on..

and what happens when AI gets "depressed?" you just task it with something? good little slave.

no they arent talking about that AI.. They are talking about regular computing with added layers or dimensions of thought on top.

this is not a dream..
They are not far off from systems that would out "smart" entire intelligence groups, entire countries.
not true AI but close enough..

cosmosmagazine.com...