Slashdot Mirror
AI Experts Boycott South Korean University Over 'Killer Robots' (bbc.com)
An anonymous reader shares a report: Leading AI experts have boycotted a South Korean university over a partnership with weapons manufacturer Hanwha Systems. More than 50 AI researchers from 30 countries signed a letter expressing concern about its plans to develop artificial intelligence for weapons. In response, the university said it would not be developing "autonomous lethal weapons." The boycott comes ahead of a UN meeting to discuss killer robots. Shin Sung-chul, president of the Korea Advanced Institute of Science and Technology (Kaist), said: "I reaffirm once again that Kaist will not conduct any research activities counter to human dignity including autonomous weapons lacking meaningful human control. Kaist is significantly aware of ethical concerns in the application of all technologies including artificial intelligence."
Asimov rolls in his grave
"...counter to human dignity"
Huh, who gets to define that?
You know what will take out robots, right ? Design one that will penetrate even the most robust faraday cage. First law of war on the battlefield... blast an EMP over enemy lines. take out central command.
There is excellent computer vision algorithms (e.g. in automotive industry) that can detect and track humans (even partially obscured) with great accuracy.
There is excellent robot technology available (just look at Boston, or your average drone manufacturer).
And there is no lack of "aim and fire" technology from the e.g. north american continent.
It is not all that difficult to assemble these pieces into a nightmarish unit. You can already now see (rather harmless soft air gun) prototypes of this on YouTube.
The university does not need to "manufacture autonomous lethal weapons", they just need to some generic AI stuff and leave the weaponization to others who probably are more than willing to do it.
This sounds like an arms race to me; if one army will obtains this technology, will the others sit around and accept it? Heck, even if all armies would collectively refrain from it, what prevents your favorite terrorist organization from doing it? It's not THAT high tech anymore.
They won't be developing "autonomous lethal weapons." They'll merely be developing autonomous technology for a company that makes lethal weapons. Nothing to see here, move along.
... then either they're incredibly naive or just stupid. Either way, I wonder if they're the sort of people to be working on a paradigm changing technology since clearly their understanding of human nature and what it will do with powerful technology is left severely wanting.
China will do it anyway.
So, people just won't develop this military tech, because peace? Got it.
No, we need to develop it, first, fastest, and best. And develop AI powered countermeasures too.
That's like almost a statistically relevant number in a population of 1,000. I guess since the world is 7 billion people though, these 50 don't really register. But wait, you say. Not all 7 billion are AI experts! Well, there are 10,000+ people that could be labeled that, so it's still not statistically relevant.
Replace the light with lead:
https://www.youtube.com/watch?...
Also in the US, we're so used to DARPA having funded cool projects. It's the last thing US-based researchers would want to boycott.
For those of you outside the US, DARPA means Defense Advanced Research Projects Agency and it falls under the purview of the Department of Defense.
I think it's funny that Google employees will circulate a petition complaining about Google working with the US security agencies, but not a peep about Google's much more robust spying on everyone. I think it's funny that people will bitch about predator and reaper "drones" with a human making every decision, but didn't notice that autonomous missiles were flying over Vietnam more than 40 years ago. I think it's funny that naive programmers and engineers say it's "wrong" to work with democratic governments on AI, knowing that the Chinese government has made a long term investment into autonomous weaponry.
Sending robots to war is better than sending people to war.
They say the following:
the university said it would not be developing "autonomous lethal weapons."
but they don't say that they aren't involved in the development of autonomous lethal weapons. They could develop all of the robotics and AI and just leave it to their partner to attach the weapons and their statement would still be correct.
Since the dawn of AI, it was funded, and still is, by the military at all universities. The technology researched and further developed by larger corporations has been both lethal and non-lethal. Usually the university does the basic research and the company adds to it.
So now some new AI hotheads are protesting developing AI (i.e. technology) that will be used in military weapons?
Technology is neither good or bad, it's how you apply it. Anyone can take the research done by a university and apply it to weapons.
I would be more scared of the research and development being done by the larger companies that you don't hear of.
They only got their degrees by constantly whining. They won't be missed!
Naively thinking AI will not be used for creating killbots is like thinking you could create the Internet and not have it used for porn.
I would say that the benefits of having an AI that can detect, react and attack targets far outweighs the downsides. The cat is already out of the bag on this one, and if research in South Korea is shut down, it doesn't mean that North Korea or China will be stopping.
Any big country is working on this research. The payoff is tremendous. Robots are cheap, and if they get obliterated, there are no body bags to send home, and who cares if Daesh chops the head off a General Dynamics robo-dog. The country that deploys UACVs will have a tactical and strategic advantage of both better reports from the battle field, no loss of morale, no long logistical chains for food, nobody needing to mix diesel fuel with poop to dispose of it. Same advantage that tanks and airplanes brought to the battlefield, making trench warfare absolutely pointless.
IS a "Killer Robot."
It fly's to its destination under its own guidance then explodes.
FAR too late to ban.
All make a big deal against it, but all nations are doing it. Russia, America, and esp china , are pouring huge bucks into this.
I prefer the "u" in honour as it seems to be missing these days.
Why haven't the AI experts banned the United States for having 1000s of super killer nuclear fusion bombs that can land anywhere on the Earth with precision?
The ban on chemical and biological weapons have been mostly effective. So I think that sets a precedent for this type of prohibition possibly working if enough countries sign on, and if sanctions against violators are both consistent and strong enough.
Not if it encourages more war because those sending the robots no longer suffer any consequences from military actions. Most modern conflicts tend to end when a populace gets fed up sending themselves/friends/parents/children to die in some far off place or if the rising costs begin to concern them, not because the conflict has been "resolved".
Welcome to the war of the thinking machines. Let the Butlerian,jihad begin.
Wasn't the nerve agent used against the Russian guy and his daughter in the UK developed in Russia after they had signed the agreement to ban chemical weapons?
Are busy design Terminator series (which would speak Chinese, Russian and not heavily accented English with a German flavor), everyone else should be snoozing?
Sure makes sense to me for them to do it, and cannot understand the umbrage?!?!?
So, now we're going to start SJW bullshit over a scientific discipline that sci-fi writers have spun horror stories out of?
Oi vey...
Chas - The one, the only.
THANK GOD!!!
That said, I have 0% faith in anyone telling me they absolutely will not weaponize AI. Anyone saying that must think we're all idiots. The first and best use for AI is replacing humans in dangerous situations, replacing them with a machine. Yeah you can do that for miners and truck drivers, but the real application will be replacing people as soldiers waging war. Period. Full stop. Okay maybe I should include law enforcement too but these days war and police work are starting to blur together at least in the US.
So, the perfect soldier. Means the perfect war. Means we get lots and lots more wars for increasingly retarded reasons including no reason of any kind. And if you happen to have played the PS4 game "Horizon: Zero Dawn" much then as you hear all these industrialists promising "we will never weaponize AI ... much" you might notice a chill running down your spine. Well my friend that is you having a premonition of Fun Things to Come.
There are a lot more ways this could go wrong than go right, and I'm even including increasingly retarded wars as "going right" just because 1) good or bad that is the product working exactly as designed, and 2) retarded wars might not end all life on Earth as we know it, especially not if there is money to be made by the weapons vendors.
But I cannot rule out weapons dealers screwing the pooch in some dead-stupid race to provide the best possible service to war-mongering, seeing as that is mostly what industrialists do when they have room to run: they four-square fuck absolutely everything up.
All the really bad things I can think of coming from weaponized AI end up with organized humans not being around. Probably won't even be an accident when (not if) it happens, probably some genuine god-damned "kill-all-*" wild card feature of the terrifying things.
Christ but we're dumb.
=^..^= all your rodent are belong to us
Why would you need a human to control the armed kill-bot, when you could just take control away from the human that ordered the deployment of the armed kill-bot?
Chas blathered:
So, now we're going to start SJW bullshit over a scientific discipline that sci-fi writers have spun horror stories out of?
It's worth noting that Isaac Asimov would definitely have identified himself as a social justice warrior, if that had been a thing when he was alive - or have you not read the Foundation series ... ?
(Posting as AC only so as not to undo prior upmods in this thread.)
--
Check out my novel ...
Not for the civilians where the robots are sent.
"Artificial Intelligence" is trying to be vastly superior to the "Natural Intelligence" at determined areas.
Once that its functionality is a success then Artificial Intelligence will be cloned to various superior robots that will defeat the dexterity of the humans.
I love that guy :)
[($)]
Sorry if you didn't have the balls to post as yourself.
The problem is, "This COULD happen with a sufficiently advanced AI", so we shouldn't pursue it AT ALL?
So, because, some day, we MIGHT eventually turn out a sufficiently advanced AI that could be dangerous, we shouldn't pursue ANY form of AI, no matter how primitive?
Sorry, that's just FUD.
Chas - The one, the only.
THANK GOD!!!