Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Preparing for Malicious Uses of AI

blog.openai.com

Post reply on HN
Print view View original

211–220 of 233 posts

Re: Preparing for Malicious Uses of AI

#211

Post by kakarot »

AJ007 wrote: post #191
DougN7 wrote: post #173

Earlier quoted context omitted.

This is the problem. WMDs take a lot of infrastructure and can be tested/inspected. AIs can be built/modified by anyone in their basement. The genie can't be put back into the bottle. It's like trying to outlaw computer viruses and hoping that will work. I don't have a solution :(

AI needs compute cycles. These are quite centralized outside of botnets. I’ve pointed this out before, but the primary commercial use cases of AI right now are malicious: mass population surveillance and behavior manipulation (ads; political and otherwise.) My biggest concern are the ML researchers who dismiss malicious use as being a distant concern while they work on projects that are malicious right now.

Homomorphic encryption + cloud services + VPN chaining. If there's one thing that won't be in shortage for the foreseeable future, it's compute cycles.

And I'm with you, the socioeconomic aspects of AI are far more important to curtail, but it doens't mean we should be ignoring the very real threat of people attacking infrastructure, sewage and power plants. And we shouldn't be waiting until after it's a problem, when it already can be seen as a certainty that it will happen eventually.

Re: Preparing for Malicious Uses of AI

#212

Post by some_puffery »

vicpara wrote: post #210

I don't think these issues have anything to do with AI but 100% with humans. AI is used everywhere where money is to be made. Let's take ads. Why do we allow ads to happen all over the place (TV, radio, internet, magazines)? Because they feel stupid, harmless, how else can we sell products, etc? Well, then when a ML system is serving you ads, it's doing it so at the right time and will target you with the bang on con…

It has at least some to do with the AI itself. Of course humans are the ultimate origin of AI, but when your system is run by algorithms created by an AI, you aren't really making the decisions anymore.

Re: Preparing for Malicious Uses of AI

#213

Post by anfilt »

tudelo wrote: post #179
anfilt wrote: post #169

Earlier quoted context omitted.

However, there is a differnce. If I dont get a job offer becuase I dont fit a humans biases of what they want. I can go somewhere else with a human that has differnt biases. The problem is with AI is how many googles are there? We will liekly end up with a handful of companies that rate you. Suddenly it does not matter where you apply for a job. If you did one thing that ruins/lowers these ratings you may not get a j…

Well, that's just part of the problem with the connected world. If you do something wrong it doesn't take machine learning or AI to find out, you just have to google it. Someone hiring you for a job will most likely do this anyways... I have had recruiters even call me out on such things. It has nothing to do with an algorithm. It could be problematic if there was only one source of subjective information, especially…

Did I say I have a problem with people making a decision based off information. Problem is these AI systems will be making decisions and if everyone is only using a handful of vendors and there software just says bad canidate or gives you a low rating. The people are just going to default to that. What it boils down to is small number of systems making decisions.

It turns into a small number making decision for many. Where as without the system it is decentralized and eveyone is making independent decisions. There also more room for nuance.

Do think conpanies are going to develop there own AI and gather datasets on people just evalute potental hires. Train it for the requirements they feel is best for there company?

No they are going to out source it. So what is likely going to happen is only small number of systems making these decisons.

Re: Preparing for Malicious Uses of AI

#214

Post by kakarot »

AndrewKemendo wrote: post #195
kakarot wrote: post #158

Earlier quoted context omitted.

Here's the crux of the issue with respect to nuclear WMDs: In order to prevent citizens and other countries from creating nuclear bombs, our government severely limits access to relevant tools and materials, and actively seeks to censor the knowledge of how to build modern nuclear devices from other parties. In order to prevent citizens and other countries from creating dangerous rogue AIs, our government ___________…

The answer is regulating data - not algorithms or hardware - the data. I'm not saying it should happen, but if you want to regulate AI then you have to regulate the data that it learns from. That means strict rules on storage, retrieval, scale, uses of, collection etc... of images, video, text, metadata, EM emissions, really anything that produces a measurable "signature" of some sort. It's how we regulate nuclear [1…

Are you serious? That kind of widespread banning of information is fundamentally incompatible with a democratic society.

It's absolutely preposterous to think that I could prevent someone from downloading and using a database of pictures, EM emissions, basically anything that a government deems dangerous, via threat of State violence.

That is as draconian and authoritarian as it gets. How do you even enforce such a law? You would need unrestricted access to every system.

Otherwise, you are only harming legitimate researchers and enthusiasts, not criminals. Criminals are still going to access and download whatever the hell they want.

How do we determine what can "help" AI and what can't? Is it the number of sources you have that should be restricted? Is there a committee that decides this? Are the proceedings open to civilian involvement? Who decides what is okay and what isn't?

Can corporations or government organizations (military) apply for this data? Why can groups of people with lots of money and infrastructure have it, but not me? How does this prevent a further segregation of socioeconomic classes?

Comparing nuclear "source material" to random data like images, video, metadata, etc is downright deceitful. If you aren't being purposefully deceitful, you really need to reflect upon exactly how such a law would be carried out, the logistics of its enforcement, and the limitations. It's a logistical nightmare and is infinitely more complex to enforce than it should be, not to mention downright anti-democratic.

Damn right it shouldn't happen.

Re: Preparing for Malicious Uses of AI

#215

Post by AndrewKemendo »

kakarot wrote: post #214
AndrewKemendo wrote: post #195

Earlier quoted context omitted.

The answer is regulating data - not algorithms or hardware - the data. I'm not saying it should happen, but if you want to regulate AI then you have to regulate the data that it learns from. That means strict rules on storage, retrieval, scale, uses of, collection etc... of images, video, text, metadata, EM emissions, really anything that produces a measurable "signature" of some sort. It's how we regulate nuclear [1…

Are you serious? That kind of widespread banning of information is fundamentally incompatible with a democratic society. It's absolutely preposterous to think that I could prevent someone from downloading and using a database of pictures, EM emissions, basically anything that a government deems dangerous, via threat of State violence. That is as draconian and authoritarian as it gets. How do you even enforce such a l…

Agreed. Again, it would be terrible for society, but if you wanted to regulate AI it would be the more effective way to go about it IMO.

There was a short story based on an implementation of this, I don't recall what it's called.

Re: Preparing for Malicious Uses of AI

#216

Post by kakarot »

AndrewKemendo wrote: post #215
kakarot wrote: post #214

Earlier quoted context omitted.

Are you serious? That kind of widespread banning of information is fundamentally incompatible with a democratic society. It's absolutely preposterous to think that I could prevent someone from downloading and using a database of pictures, EM emissions, basically anything that a government deems dangerous, via threat of State violence. That is as draconian and authoritarian as it gets. How do you even enforce such a l…

Agreed. Again, it would be terrible for society, but if you wanted to regulate AI it would be the more effective way to go about it IMO. There was a short story based on an implementation of this, I don't recall what it's called.

It's not at all the most effective. It's only the most effective if you severely limit what factors into that.

Re: Preparing for Malicious Uses of AI

#217

Post by basementcat »

Aco- wrote: post #131
nerdponx wrote: post #56

Earlier quoted context omitted.

already accept it's happening and inevitable I've almost given up. I'm pretty much despondent at this point. There is a booth on my way to work where they give out free Lays chips in exchange for your photo. The gimmick is that it's like a photo booth. I asked the attendant to read the privacy policy, and it's frightening, to say the least. I have no doubt that they are contracting with some AI consultant to be able…

You say MTA, so is this booth in NYC? Where exactly is this Lay's spying machine so I can steer clear.

Those of us with the munchies also want to know the location of this snack dispensing machine.

Re: Preparing for Malicious Uses of AI

#218

Post by PeterisP »

VMG wrote: post #19
zucchini_head wrote: post #16

I find it quite funny, rather intruiging, that we seem to have gone full circle on trusted sources of information. Historically, a face-to-face meeting was considered as the ultimate legitimate and trustworthy way. Not story or rumors or witnessing, since the courts say people can be "decieved", "traumatised", etc. Then came microphones, cameras, CCTV in the 20th century, and then they became the ultimate trusted sou…

There is and will be a place for strong crypto for trust, however probably not on standard commodity hardware.

Crypto works for trust of computing systems - it can make some claims about the transformations of the captured data or about the integrity of the software handling these transformations, but it fundamentally can't make solid claims about reality that's supposedly being captured. You can't put crypto between reality and a sensor.

At best, crypto can give you a statement like "something possessing the particular secret X claims to have sensed this data at this point of time" combined with "there's a device with secret X that has it's software/firmware integrity verified and signed by entity Y". Crypto can't ensure that the secrets on the device aren't actually leaked by the manufacturer to enable "verifying" of arbitrary data outside of that device, and there's always the option to simply ensure that the sensor "sees" what you want; a camera and all the crypto on it can't tell whether it's pointed at a real event, at a staged event, or at a sophisticated optical device projecting arbitrary photoshopped data.

Re: Preparing for Malicious Uses of AI

#219

Post by letlambda »

Mizza wrote: post #60
letlambda wrote: post #50

Earlier quoted context omitted.

How is your concern different than what they are calling political security (~1/3 of the paper)? >Political security. The use of AI to automate tasks involved in surveillance (e.g. analysing mass-collected data), persuasion (e.g. creating targeted propaganda), and deception (e.g. manipulating videos) may expand threats associated with privacy invasion and social manipulation. We also expect novel attacks that take ad…

I guess it's the difference between what they're calling "authoritarian states", the mustache-twirling dictators of the 20th century targeting their political opponents (a fetish in papers like this), and the _new_ form of corporate-driven mass surveillance and resulting fascism that is emerging. There is no need for an evil central power targeting threats to itself, but instead we have seen/are seeing the rise perva…

I haven't read the paper, but I think:

>undermine the ability of democracies to sustain truthful public debates.

Is coded language for exactly what you're talking about, in the same way that dictators is coded language for geopolitical enemies.

Re: Preparing for Malicious Uses of AI

#220

Post by lern_too_spel »

canaanical wrote: post #142
banderman wrote: post #90

Earlier quoted context omitted.

It's the same as the Microsoft "How old are you?" and Google "What painting do you look like?" apps. People don't think about what's going on behind something that seems trivial.

Anecdotally, Google Opinion Rewards first question was whether or not I supported the UN. I immediately deleted the app.

Somebody, probably a political organization, ran an opinion poll on the UN and got anonymized aggregated results. What part of that made you delete the app? If you get a political phone survey, will you unplug your phone?
Post reply on HN