Without paywall:
Couple things:
AI is the least efficient way to do literally anything. It’s just faster at a few very specific things.
AI does not lean toward quality. That is not a metric it can comprehend. It is not a pressure it can respond to. Predictably, its quality has and will always be horseshit.
Finally, who looks at Hitler and thinks the problem was poor optimization?
who looks at Hitler and thinks the problem was poor optimization?
Neo nazis in middle management
It says something about our current world that, having read the headline and not knowing who this guy is, I expected him to a booster of both AI and Nazism. It’s a relief to find he means this as a criticism.
For a moment I thought it was onion
Ok, i appreciate what he wants to say, but he has fully bought into the delusion that an LLM gets somewhat close to an AGI, including the “LLMs could wipe out humanity” crap that Musk/Altman/Amodei are pushing to make their models seem more capable than they will ever be. The only way an LLM can be a danger to humanity is when it gets integrated into automated killchains including a nuclear arsenal, and i hope noone is stupid enough to hook up the nuclear football to a statistical model that is guaranteed to have malfunctions.
It’s not just about current LLMs, though. LLMs are being made to work 24/7 on the next generation of models (and not just LLMs), which may be quite novel or just more effective. The next generation might be the one to worry about… Or the generation after that.
LLMs can be used like an army of monkeys with typewriters, with evals guaranteeing progress. It’s inefficient, but effective.
Assuming it is possible to make progress (and I think it is), the logical conclusion is that progress will be made, and AGI and ASI will come
it gets integrated into automated killchains including a nuclear arsenal, and i hope noone is stupid enough to
Have you ever met a human? 🤣
The current landscape is the wet dream of nazis. It’s gonna be so efficient to dissapear people. They even put their own information online. Silly ones.
What a choice of words… 🤦 Deliberate? “Better”? How about “worse” instead?
I’d prefer someone worse at killing people
I would appreciate gen AI being even better at killing itself.
Anyone help me get past the German paywall?
Brother America’s been doing it before Ai, it was just marketed differently.
I don’t think it’s wrong to look at any technology and decide whether it’s worth perusing based on “how excited would hitler have been to use this”
Stuff like databases gets a pass, that’s just an essential thing. Stuff like AI and Flock obviously crosses the line into “could easily be used for genocide at scales never seen before”
Depends on the purpose of the database.
Yes, it’s the what not the how databases work
I have not read the article yet. Just popped in to say the headline alone strikes dread in the very core of my soul.
Well, that’s not hard, Hitler was smashed 24-7 and he was demonstrably crazy
Leave it to AI to find a more efficient way to shoot yourself.
Oh, he’s talking about the other thing…
They don’t need another selling point!
Stuart Russell: ‘What Hitler Did, AI Could Do Faster, Better and More Efficiently’
Depending on how much we integrate it into all our shit I fear it could also do it imperceptibly.
How could A.I. butchery be “better” ? It is still butchery all the same.
Just straight up done listening to old white shits completely.
I think you’ve missed the point a bit. It’s “better” only from the Nazis’ point of view. Stuart Russell is not saying that it would be better. He thinks it would be worse, because people or AIs being better at Nazism is worse for the world overall. He’s saying AI is a threat to the world because right now it’s in the hands of people (many of them fascists) who ignore the dangers.
Here’s the whole thing without paywall:
My old shits usually remain brown 😉 Mine never talk back though!
You can ignore it. It’s just a circle jerk of hate. It would make more sense posted on /fuck_ai but that community leaks on the daily.
