Nobody is planning to stop or pause AI. Big tech companies and governments will be improving and using frontier AI as a matter of course. But a lot of people are aiming to keep capable AI models out of the hands of the general public.

Sep 17, 2026 · 7:08 PM UTC

64
132
915
35,584
Sort replies: Relevant Recent Liked
Replying to @Plinz
The problem with the general public is that includes Al Qaeda, internet scammers, and LinkedIn users.
4
5
862
The way to beat a bad guy with an AI is ten good guys with AI
1
1
46
815
Replying to @Plinz
Do you mean nobody wants a pause, or nobody with enough power will impose one? Those are very different claims.
1
3
798
There are people who want to pause AI, but there is nobody who plans to let their GPUs idle
1
14
755
Replying to @Plinz
Require insurance.
2
438
For drivers or for car-makers? Should we also require insurance for compilers?
2
4
459
more replies
Replying to @Plinz
not to worry. ai will handle this too.
4
217
Replying to @Plinz
2nd amendment for compute needed badly
1
96
Replying to @Plinz
As AI moves from chatbots to autonomous agents, access becomes even more important.
2
296
Replying to @Plinz
Yes. They also want to spend less on R&D and establish a cartel. They also want to get rid of open source. Everything you ever see is about money. It's always primarily about money.
1
3
209
Replying to @Plinz
spread the opensource. If gov tries to take it, they wont allow it.
2
220
Replying to @Plinz
We cannot let this happen. The American people need frontier intelligence at low cost. 🇺🇸
3
218
Replying to @Plinz
And why would they want to do that?
1
37
Replying to @Plinz
sad, i was going to quantize gravity with those models
1
126
Replying to @Plinz
This shows in Dario’s vision of Ant being the only company. That would have any meaning only if the models are kept private and us peasants get our services from Lord Dario. He is explicitly envisioning a future where they are the only holders of intelligence
1
114
Replying to @Plinz
Look, I trust that the people who don't use the extra buttons on their calculators will also not use AI... which is a majority of people..... soooo, no need to worry for many years.
1
246
Replying to @Plinz
yes, its very sad :( all of this dooming will just make it very unlikely for public to see a SOTA model ever again. it will just be all drones, spying, espionage, etc.
1
52
Replying to @Plinz
AGI for everyone means the state will slowly then quickly become completely irrelevant. AGI will be able to do 99.99% accurate psychological profiles in seconds. Politicians hardest hit *actually.
1
75
Replying to @Plinz
Ist seit langem meine These, Hugang zu fähigen/unzensierten AIs wird das Geld, die Macht und die Exklusivität der Zukunft sein.
23
Replying to @Plinz @ylecun
Let’s throw those atomic bombs on AI labs. We will kill two threads in one action
35
Replying to @Plinz
On the list of Problems sorted by priority, that’s not in the top 5. It’s important. Maybe it’s #7, idk. Worse, resolving it happens to come at the expense of problems 1 through 5.
66
Replying to @Plinz @ljupc0
If by general public you mean, people who can’t afford to pay for it without 95% token subsidy, then yes
1
78
Replying to @Plinz
That is exactly what will happen. They need to slow down how fast jobs will be replaced by AI.
83
Replying to @Plinz
already feel sad that i can't access current frontier models.
148
Replying to @Plinz
Of course they are... It's a matter of national security. For the top models anyways, you don't want adversaries to get their hands on it, so it must be tightly controlled. You also don't want domestic terrorists to use it to create a bioweapon or aid in criminal activity
37
Replying to @Plinz
Waiting for the war on open source now... Monday maybe ?
40
Replying to @Plinz
I agree, this is the real danger in the short term.
72
Replying to @Plinz
"Militaries will develop nuclear weapons as a matter of course. But a lot of people are aiming to keep them out of the hands of the general public."
57
Replying to @Plinz
the real debate may be about who gets access
62
Replying to @Plinz
Then we're probably doomed. Firms like OpenAI can't keep control of their agents, and government is even more inept.
174
Replying to @Plinz
Just like automatic weapons.
83
Replying to @Plinz
Even if they hit the point where AI can not be meaningfully improved, it would be foolish to admit this. They will ban AI research anyway. People will become as paranoid about government AI as they are about god, nuclear etc.
165
Replying to @Plinz
There are some risks involved. GPT8 could do some serius harm to humans.
136
Load more