"Ethical AI exists" is very much so "obeying in advance."

I'm not a huge fan of "don't obey in advance," as it implies that there is a correct and useful point at which one should obey fascists. "Fuck you, make me" or even just "fuck you" work much better.

But goddamn it, "ethical AI" *is* obedience. It is acquiescence. It's the narrative refuge of quisling appeasers. It is an absolutely *vapid* intellectual stance that does nothing but muddies the waters when it comes to the fascist project of enclosing culture with AI.

Reasonable person: AI is bad and here's why.
Professional sealions: But wait, ETHICAL AI.

I don't say this kind of thing often, but I'm going to ask for a favor here in a pretty rude way.

If you feel compelled to tell anyone about "ethical AI" or to start soapboxing about "AI ethics," please please please shut the fuck up. Just, seriously. If you aren't going to help, at least don't jump in and start causing even more problems.

This isn't about any one person or any group of people, it's just this whole bullshit argument that I keep seeing over. and over. and over again. You just don't gotta do that shit. You don't have to make excuses for AI that way.

Gonna end the thread by admitting, as ever, that I'm just one woman out here on Satan's own internet. I can be wrong. If you think I'm wrong, and you really need to get that off your chest, at fucking *least* show up with something that hasn't been said a thousand times before. I promise, I'm terminally online. I've seen all the typical "ethical AI" arguments before.

Maybe bring some evidence with you. For a lark.

@xgranade
I do see uses for it that would be ethical, but not profitable. And unless they can be extremely profitable with it constantly, it won’t be maintained. So at this point in time, with all we are facing, and no sustainable infrastructure to support all the water, energy, planet, climate, & people that will be harmed, I see no way it can be ethical at this time, weighing all the costs.
It’s grossly premature to market.

@xgranade ethical AI is actually just sparkling carefully applied machine learning

The whole point of the AI salescon was to make people accept under developed poorly tested machine learning models in deployment by borrowing from fictional future credibility. All the while they cast standard human and statistical systems as fundamentally stagnant and immutable. It's an obvious con job.

@xgranade it's so frustrating because the idea of ethical AI is legitimate, but it's absolutely not what people are talking about when they say "ethical AI."

Train your own neural network model on your own dataset for a specific purpose? That's both cool *and* fascinating. Basically no one is doing that, though, and the ones who are getting drowned out.

@xgranade I had an idea to make a game where the rendering is done via a neural network trained on rasterized screenshots of that same game with the intention that it would create a dream-like aesthetic for artistic purposes.

Now the idea of telling people "I made a game where the graphics are fully generated by a neural network" would get me crucified even though that's probably the most valid application of the mathematics.

@xgranade
AI ethics? You mean turning it off and putting those behind it in jail?

@xgranade

I hate that AI ethics lumps together people who get all Roko's Basilisk about LLMs and people who do things like study how racial bias in training data gets amplified or write papers about how there is no even kind of ok way to use any AI system for targeting weapons

@gbargoud @xgranade

"people who do things like study how racial bias in training data gets amplified or write papers about how there is no even kind of ok way to use any AI system for targeting weapons"

It's really hard to even talk about this kind of study without buying into the lie. :(

@xgranade

Me: "It's fucking not even AI"

@xgranade I am assuming you are talking about genAI. If that's what you mean, I agree. There could have been ethical genAI, but that's not what happened and you can't undo how those models were built. Everything downstream is tainted now, forever.

@xgranade The reason they are pushing AI so hard is that it obliterates critical thinking.

https://youtu.be/KhBsHoiiorM?si=yYD1b6izltMD73FW

The very reason Trump signed an Executive Order to ban state legislation on AI.

MIT Researchers Discover How AI Changes the Way Your Brain Works

YouTube
@xgranade I can’t help but think if I had the ability to take all information and control it forever I would be a very bad person doing very bad things and anybody trying to do good with that is only making me stronger? I don’t know why this is hard