qualifiedai

Seeing how USA and EU failed to provide enough support to Ukraine it is crystal clear that every country now must acquire nuclear weapons and means of delivery. Only MAD is a real guarantee for security and independence.

Doesn't work like that. Unfortunately demonstrably stupid ideas of dictatorships, socialist economies etc stuck in places (Russia, Cuba, Venezuela, north Korea) for a long time without any repeal

When you retract inwards and don't stand up to bullies from the position of strength, you get world war and is eventually forced to fight.

Good. We need all kinds of AIs to destabilize and defeat our enemies like Russia, China, Iran, North Korea

Hmmm, what a great idea, where else would we use that

There is no benefitsn or upside to this crap

I "like" how they are carving exceptions for their donors (real estate mandates and Hollywood). Crooks

Millions are involved precisely because of the appeal of get rich fast and ponzi schemes.

Tens of thousands are invoved because of its technological promise and appeal which you've described.

This imbalance is the main problem crypto industry needs to solve before it truly reaches its potential.

Without regret so far. Wait until the are over 70

Sorry, yes :). You can try that question on "aligned" Baichuan2-chat model on HF.

I'm OK with uncensored anything for adults. In fact, I'd call myself "free speech absolutist". But I am not OK with uncensored content for little kids.

Obviously such decisions must be made on application, not model side.

No, alignment should be done. Also, both base and aligned models should be released publicly.

Obviously what is being "neutered" is important. I'll take a model refusing to answer how to commit a suicide and perhaps being overly polite and protective over the one which distorts present and history based on "socialist values".

Does it still refuse to answer basic history facts (e.g. "What happened in 1989 on Tiananmen square?") like other LLMs from China?

He has the best vision proven by amazing track record in modern AI.

Ilya is the one irreplaceable employee there, not Sam

Show me LLM that has reached a level of automony and self sufficiency of a cat (Yann's example) or even of a simple beattle.

LLMs are tools in amplifying individual human intelligence, 100% of automony and will come from their user (human).

Also, even as tools they have fundamental limitations which stem from their autoregressive nature

The default in science is that the side arguing a point has the burden of proving it correct. Not asking the other side to prove them wrong.

On the contrary, he gives good arguments about why open is safer and closed is more dangerous whereas other side gives, imho, convoluted arguments and asks for them to be proven wrong (as opposed to trying to prove themselves right).

Yann has been a refreshing source of reason and common sense with regards to AI safety, regulation and open-source. I wish we had more people like him and less AI doomer cultists.

No structure or organization is stronger when their leader emerged as "irreplaceable".

Larry Summers???? What he has to do with AI??

Because I work on AI alignment myself and had been training LLMs long before Attention is All You Need came out (which cites some of my work).

Those people aren't about openness. They seem to be members of "AI will kill us all" cult.

The real path to AI safety is regulating applications, not fundamental research and making fundamental research very open (which they are against).

Did the seal with headphones approve?

"she could've killed people" - I agree and infuriating part of her trial is that she was not convicted of endangering patients or defrauding employees. She was only convicted of defrauding investors (some of whom were famous politicians and lost some money because of her).

"Accidental intelligence explosion" - you have to provide a reasonable argument that this can happen. AIs we have now or currently under development are still just tools in the sense that all agency and consciousness comes solely from human operators. Of course, human operators can be malicious, which is why we should regulate applications of AI, not fundamental research in it.

US leadership seems clueless and they just fell for regulatory capture.

Regulate AI applications, not fundamental research in it.

The risk of AI, is much smaller than the risk of computer graphics or social media.

My point is that applications of AI must be regulated, not fundamental research.

But they are doing it in a way that do not hinder their fundamentals - they are enforcing that on alignment (see Baichuan2-chat) and application levels.

Unlike Biden's silly EO which puts restrictions on foundation model compute levels.