It’s also about distraction. The main point of the letter and the campaign behind it is slight-of-hand; to get the media obsessing over hypothetical concerns about hypothetical future AIs rather than talking about the actual concerns around current LLMs. They don’t want the media talking about the danger of deepfaked videos, floods of generated disinformation, floods of generated scams, deepfaked audio scams, and on and on, so they dangle Skynet in front of them and watch the majority of the media gladly obsess over our Terminator-themed future because that’s more exciting and generates more clicks than talking about things like the flood of fake news that is going to dominate every democratic election in the world from now on. Because these LLM creators would much rather see regulation of future products they don’t have any idea how to build (and , even better, maybe that regulation can even entrench their own position) than regulation of what they’re currently, actually doing.
I’m going to need a legal framework to be able to DMCA any comments I see online in case they were created with an AI trained on Sara Silverman’s books
Since I don’t think this analogy works, you shouldn’t stop there, but actually explain how the world would look like if everyone had access to AI technology (advanced enough to be comparable to a nuke), vs how it would look like if only a small elite had access to it.
We could all do our taxes for free. Fix grammatical errors. Have a pocket legal, medical advice. A niche hobby advisor. Pocket professor. A form completion tool. All in one assistant especially for people who might not know how to navigate a lot of tasks in life. Or we could ban it because I fear maybe someone will use it to make memes. Lots of lazy articles convinced me the AI sky is falling
Okay, well, if everyone had access to an AGI, anyone could design and distribute a pathogen that could wipe out a significant portion of the population. Then again, you’d have the collective force of everyone else’s AI countering that plot.
I think that putting that kind of power into the hands of everyone shouldnt be done lightly.
You still can’t manufacture it. Your comparision with nukes is actually a good example: The basic knowledge how a nuke works is out there, yet most people struggle in refining weapon-grade plutonium.
You can’t type “How do I make a pathogen to wipe out a city” into a book. A sufficiently advanced and aligned AI will, however, answer that question with a detailed list of production steps, resource requirements and timeline.
this requires special materials like enzymes and such. It would much easier to restrict access to those. Now true this godlike ai could go back to show you how to make all the base stuff but you need equipment for this like centrifuges and you will need special media. Its like the ai telling you how to make a nuke really. Yeah it could star you off with bronze age metal smithing and you could work your way up to the modern materials you would need but realistically you won’t be able to do it (assuming again you restrict certain materials)
Then I’ll ask AI how to obtain plutonium, checkmate.
But by that point I might just ask the all-knowing AI how I can achieve what I want to with the nuke and cut out the radioactive middle man. Unless the AI tells me to build a nuke, then it’s nuke time anyway.
The point I was trying to make is, all the information about viruses and nuclear bombs are already readily available. AI doing the googling for you will not have an actual impact, especially considering what else you’ll need to make it all work.
I would assume you get the fear of AI from the news media. Understandable, they have a vested interest in keeping you afraid. AI is gonna steal their ad revenue, when you won’t have to visit their shitty websites anymore.
So, the actionable suggestions from this article are: reduce competition and ban open source.
I guess what it is really about, is using fear to make sure AI remains in the hands of a few…
It’s also about distraction. The main point of the letter and the campaign behind it is slight-of-hand; to get the media obsessing over hypothetical concerns about hypothetical future AIs rather than talking about the actual concerns around current LLMs. They don’t want the media talking about the danger of deepfaked videos, floods of generated disinformation, floods of generated scams, deepfaked audio scams, and on and on, so they dangle Skynet in front of them and watch the majority of the media gladly obsess over our Terminator-themed future because that’s more exciting and generates more clicks than talking about things like the flood of fake news that is going to dominate every democratic election in the world from now on. Because these LLM creators would much rather see regulation of future products they don’t have any idea how to build (and , even better, maybe that regulation can even entrench their own position) than regulation of what they’re currently, actually doing.
I’m going to need a legal framework to be able to DMCA any comments I see online in case they were created with an AI trained on Sara Silverman’s books
But… shouldn’t it? I mean, if everyone had a nuke, the world would look a whole lot different
Since I don’t think this analogy works, you shouldn’t stop there, but actually explain how the world would look like if everyone had access to AI technology (advanced enough to be comparable to a nuke), vs how it would look like if only a small elite had access to it.
We could all do our taxes for free. Fix grammatical errors. Have a pocket legal, medical advice. A niche hobby advisor. Pocket professor. A form completion tool. All in one assistant especially for people who might not know how to navigate a lot of tasks in life. Or we could ban it because I fear maybe someone will use it to make memes. Lots of lazy articles convinced me the AI sky is falling
Okay, well, if everyone had access to an AGI, anyone could design and distribute a pathogen that could wipe out a significant portion of the population. Then again, you’d have the collective force of everyone else’s AI countering that plot.
I think that putting that kind of power into the hands of everyone shouldnt be done lightly.
You still can’t manufacture it. Your comparision with nukes is actually a good example: The basic knowledge how a nuke works is out there, yet most people struggle in refining weapon-grade plutonium.
Knowledge is only one part in doing something.
I would say the risk of having AI be limited to the ruling elite is worse, though - because there wouldn’t be everyone else’s AI to counter them.
And if AI is limited to a few, those few WILL become the new ruling elite.
And people would be less likely to identify what AI can and can’t do if we convince ourselves to limit our access to it.
People are already incompetent enough at this when there’s a disclaimer in front of their faces warning about gpt.
We’re seeing responses even in this thread conflating AGI with LLMs. People at large are too fucking stupid to be trusted with this kind of thing
Are we back to freaking out about the anarchists cookbook
There are papers online on how to design viruses. Now to get funding for a lab and staff, because this is nothing like Breaking Bad.
Your brain is an (NA)GI
You’re just gonna print the pathogens with the pathogen printer? You understand that getting the information doesn’t mean you’re able to produce it.
I need an article on how a 3d printer can be used to print an underground chemistry lab to produce these weapons grade pathogens
That’s the thing though: a sufficiently advanced intelligence will know how. You don’t have to.
I know how to build a barn. Doesn’t mean I can do it by myself with no tools or materials.
Turns out that building and operating a lab that can churn out bespoke pathogens is actually even more difficult and expensive than that.
Since when does AI translate to being able to create bacteria and stuff?
If having the information on how to do so was enough to create pathogens, we should already have been wiped out because of books and libraries.
You can’t type “How do I make a pathogen to wipe out a city” into a book. A sufficiently advanced and aligned AI will, however, answer that question with a detailed list of production steps, resource requirements and timeline.
Oog, what if by making this fire, it burns down the forest?
Well that did happen to be fair.
Too soon
Right. So, the actual danger here is… Search engines?
this requires special materials like enzymes and such. It would much easier to restrict access to those. Now true this godlike ai could go back to show you how to make all the base stuff but you need equipment for this like centrifuges and you will need special media. Its like the ai telling you how to make a nuke really. Yeah it could star you off with bronze age metal smithing and you could work your way up to the modern materials you would need but realistically you won’t be able to do it (assuming again you restrict certain materials)
Have you heard about this thing called the internet?
Let’s assume your hypothetical here isnt bonkers: How, exactly, do you propose limiting people’s access to linear algebra?
You can google how to make a nuke. Of course, you’re gonna get your hands on the plutonium, which is something even countries struggle with.
Then I’ll ask AI how to obtain plutonium, checkmate.
But by that point I might just ask the all-knowing AI how I can achieve what I want to with the nuke and cut out the radioactive middle man. Unless the AI tells me to build a nuke, then it’s nuke time anyway.
The point I was trying to make is, all the information about viruses and nuclear bombs are already readily available. AI doing the googling for you will not have an actual impact, especially considering what else you’ll need to make it all work.
I would assume you get the fear of AI from the news media. Understandable, they have a vested interest in keeping you afraid. AI is gonna steal their ad revenue, when you won’t have to visit their shitty websites anymore.
Don’t worry, I was just meming off of your point. Look my other comment in this thread.