Ai in general is an area of active research. The AIs that will bring about the end of the world do not exist — they are not simply in need of an implementation.
Right now ai is driven by the collective effort of many research entities. Universities and private businesses. Ai research in those places can be shut down very effectively with regulation. That would probably be enough to buy us centuries because progress in ai depends heavily on this open community of research.
Making the use of high level signal processing illegal for business would further delay ai because there would be no incentive for anyone to develop models for profit. We audit taxes and lots of other things, auditing signal processing is not obviously impossible.
As you mention, cloud computing is a thing. Cloud computing will likely be the nucleation point from which the first agi springs. It should be shut down. Through regulation or through the dismantling of the internet it should be shut down. Shutting it down would buy us lots of valuable time.
ICs are not something you can make in your back yard. Controlling the compute density and energy efficiency or any other aspect of ICs is trivial because they require huge factories to manufacture. Huge amounts of space, energy and money are required to run those fabs. And there aren’t very many of them. Therefore they are highly susceptible to regulation and public oversight. We could stop the production of ICs that are too powerful.
So, here you are in your basement. You are trying to train an AGI. You are stuck with old hardware that you somehow bought on the black market. You are risking your life to do it because it’s highly illegal. You don’t have any investment, legal or otherwise, because it’s highly illegal and also stands to make no profit for anyone. You don’t have access to cloud computing. The training is very slow. It is made even slower by the fact that you have to slowly charge a gigantic battery array before each training because without batteries the spike in wattage would be inexplicable for a house in the suburbs and probably trigger a visit from the power authority. The probability that anyone would succeed in this scenario is not zero, but it’s very low. Very, very low. And I think it would save the world.
I’ve already read the blog post and watched the video. I’m not just some idiot who reads a few headlines and then comes to this conclusion. I found lex’s conversation to be frustrating beyond description. These people simply don’t understand the consequences of this technology. They think that an ethics board will help in some way. They are completely missing the point.
“What’s going to stop bad things from happening?”
“The charter”
Good fucking god.
The burden is on your camp to prove it’s claims, not mine. If the burden of proof is on anyone, then it should be on the people who claim it’s safe because we stand to lose everything if it isn’t. If my lot is wrong, we lose nothing.
Unwelcome huh? I’ve been here longer than you have.
That’s not correct. What you are saying is that there is no plausible organized effort that could stop or slow the creation of signal processing models that will have pronounced negative impacts. The error is on two levels: you are using too much analogy with other technologies. And you are writing off the possibility of stopping ai when it’s still not clear that it can’t be stopped.
This isn’t something that can be built and tested in isolation like other things we are familiar with. Training these models is not an exact science. Nothing about ai is an exact science. Progress only comes with trial and error. And each trial requires huge compute resources; at least for the most capable and dangerous models. It can’t be done in your basement. Not without significant effort and drawing attention to yourself. Could we sense whenever someone was trying to do it? Could we form a global coalition to stop every attempt? That brings us to the next thing.
What you are doing is the following: we are both in a car that is about to roll off a cliff. I propose that we try pressing the brakes. You respond by saying that, geez it looks like we probably wouldn’t stop in time — we are going awfully fast and it probably wouldn’t work to press the brakes so why even try? Let’s just brace our heads and hope the impact doesn’t kill us.
Obviously the better thing to do is to try and press the brakes. Even if you aren’t sure if you can stop in time.
Here’s a summary: we are aware of the fact that this model will harm society and we are releasing it anyway. We are fiddling around with the way it’s released in an attempt to absolve ourselves of blame while simultaneously collecting the profit in the form of a juicy acquisition.
The net result of these advanced forms of signal processing will be negative. Nobody has come forward to prove that they will benefit society on the whole or even that they are safe. But anyone who raises concern is shouted down and called names like “alarmist” and “Luddite.”
These companies are playing with fire, and the whole world stands to be burned. Wake the fuck up.