Forgot your password?
typodupeerror

Comment Can you hear the sound of the AI bubble popping? (Score 2) 23

This is the beginning of the end for the AI bubble. And there's reasonable concern that may lead to recession if it's true the AI gold rush is what has been propping up economies in the face of all the other increasing political and economic turbulence.

AI is hitting a different wall than most people have been talking about. Before any AGI or "technological singularity", government is stepping in (for whatever motives/justifications) and restricting access.

New models can still come, but the most advanced will no longer be for the mass market. We're entering a new class-based regime where you have to be worthy to qualify for the most powerful new capabilities. And government is deciding who is worthy. So the new stuff will only be available to a privileged few.

To be fair, Anthropic and others (plus even OpenAI in the past) have been calling for society to create guardrails. They say the tech mega-corps have seemingly irresistible financial motives to rush AI forward with little regard for consequence. There are some compelling arguments there. And this is one initial answer to that wish. (Though perhaps authoritarian power is not the best answer.)

The bubble bursts because this is the end of massive releases for the most breathtakingly capable new LLM models. AI companies can no longer juice their investor appeal by throwing more $$$Billions at larger and larger models because now only a few can use them.

So investors will pull back. The gold rush ends and the "irrational exuberance" bubble pops. Investors will start to finally ask the questions they should have been about how supportable all these future contracts are for data centers, power, RAM, GPUs, etc. They can see the end of unlimited growh for AI. And the sell-off of companies without convincing answers will start.

Eventually, like in the DotCom bust, the underlying real value will remain and rise again after the downturn. But there's likely some hurt between here and there.

Comment Re: Antropic literally asked for this (Score 1) 40

Well they didn't quite ask for this, but they rightly are looking for some rational government regulation on the whole industry. When they say "You shouldn't trust us to restrict ourselves, because market competition will steamroll any company who holds back when the others don't", it's true.

Hopefully, this will teach him why many are sceptical of asking governments to regulate against innovation.

Healthy skepticism of government is always worthwhile. Until it becomes a habit of outright suspicion and automatic disbelief. Because "the government" literally is us. It's how we do things as a society. If it's crap then it's our fault for voting for it. It's our fault for not listening to and respecting each other enough to reach consensus and make it get better.

Comment Re: Be careful what you wish for (Score 2) 56

And as pointed out in a Medium article (Paywalled, sorry), the whole Fable/Mythos release restriction is also the beginning of a class-tiered "You aren't worthy of this" regime.

https://medium.com/@ignacio.de.gregorio.noblejas/todays-the-start-of-a-new-ai-era-you-aren-t-invited-to-e0172b6ac3b8

Welcome to the Brave New World

Comment Re: Be careful what you wish for (Score 2) 56

Thank you, this is exactly what I came here for:

If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers.

It would however represent exactly the type of "please regulate the industry; we shouldn't possibly be trusted" response Anthropic advocates. Just a typically Trumpian-style instance of it.

To wit:

the order suggests all "non-Americans" would be restricted from using Anthropic's latest models, including those based in the U.S. "This means you should expect to have to prove your citizenship to use Anthropic models,"

Comment Re: Slashdot: "Panic !" Also Slashdot: "Don't Pani (Score 1) 240

The referenced Washington Post article is based on US government statistics, and if you believe those statistics, I have a bridge to sell you

Certainly it's hard to know whose stats to believe. But just automatically/angrily dismissing official agency statistics as MAGA-fake-news is just as dangerous as the *Rump admin's attempts at political meddling and labeling everything they disagree with as fake news. It lets them win at the more important game of destroying all confidence in institutions.and sources of truth.

Instead the better approach is to track down trustworthy sources to refute those statistics, if they indeed have been biased. Publicize the conflict widely. Investigate and rationally weigh what is true or not..

And honestly and respectfully engage with your friends who may disagree. A large fraction of America VOTED for this shit sandwich we're all having to eat.. The only way to fix this is to change people's minds.

Most people tend to think that's an impossible and naive approach. It certainly is hard and slow. But even when they're dead wrong, everyone has the right to their own beliefs and opinions. You can't change someone's mind before you recognize their inherent human dignity and accept their right to be where they are. You need to openly explore with them WHY they believe what they do. Then you can at least try to honestly discuss the issues.

To quote Ted Lasso (not in fact Walt Whitman): "Be curious, not judgmental."

And everyone needs to remember at the polls.

Comment Re: The short answer is: no (Score 1) 141

Thank you! I came here to point out that allowing states to exempt themselves from Daylight Savings Time is absolutely NOT what this is about. That would, as you say, put them on legitimate solar time year round.

I for one would vastly prefer that. Let businesses and agencies set the hours that work for them and their customers. But make them do that themselves, instead of asking government to change the clocks and force everyone by default.

But in our post-truth, post-reality world, this is how things are done.

Comment Re:Maybe. Or maybe not. (Score 2) 77

Thank you, came here for exactly this!

It's immediately striking there are no independent validations of any of these self-interested claims by a commercial developer. It looks like an oversize bacckpack leaf blower. Nothing that seems like it could effectively cover a large area or a home interior. What, it's going to BLOW this through our HVAC ducts??

I'd also demand some solid safety certification of those claims of "harmless to pets and humans" . . .

And finally, yeah - the New York Post?? Sheesh! ChatGPT summarized it well:

The New York Post has a low-to-mixed journalistic quality reputation: influential, fast, and widely read, but known for tabloid sensationalism, right-leaning framing, aggressive headlines, and uneven factual reliability.

Comment Link the TFA! Please (Score 2) 25

Come on, I know Slashdot is far from real journalism, but this is an important news item. From a source (credited, at least by name in text) I hadn't encountered, but that appears to have valuable credentials. To find TFA I had to first search up The Conversation's site, then search there for the article. The Slashdot post doesn't even include enough of the TFA title to easily search it.

Yet even an outrage piece about a fucking Pokemon Go tournament leads with a link to TFA. Entertainment above journalism.

US government ramps up mass surveillance with help of AI tech, data brokers – and your apps and devices

You're welcome

Slashdot Top Deals

The only perfect science is hind-sight.

Working...