Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Identity verification on Claude

support.claude.com

Post reply on HN
Print view View original

661–670 of 765 posts

Re: Identity verification on Claude

#661

Post by sneak »

christoph wrote: post #169
samename wrote: post #153

Earlier quoted context omitted.

Thank you. This is an expansion of surveillance. Most concerning to me is how few people understand government's access to this data. People treat AI like a diary, and the government can request that data at anytime.

Indeed - Persona is backed by Founders Fund - also linked to lovely companies like Palantir, Flock & Anduril. Some people still (claim they) can’t see the massive dragnet swirling around all of society right now.

Make sure you write it fully qualified:

Flock (YC S17)

Flock is a Y Combinator company.

Re: Identity verification on Claude

#662

Post by sneak »

mad_tortoise wrote: post #189
jsyang00 wrote: post #177

Earlier quoted context omitted.

> Pointless CYA They are literally being forced to do this by the government. The counter-ask if for them to self immolate in order to take a principled stand, as access to all their models is banned.

Or move to Europe/China/insert lovely tax haven here. They don't have to base their operations in the USA, they CHOOSE to. I will cancel my subscription because of this, and happily use DeepSeek/mimo or whatever else comes for a fraction of the cost. China won't and mostly can't do anything with my data, the US government can and certainly will. I'll pick my poison thanks and choose Chinese surveillance.

There is no way for Anthropic to move and remain competitive and useful. Their staff are (primarily) in the US. Thousands of people with mortgages, marriages, kids in school, recurring medical appointments with their trusted providers, people who speak only English. This idea that US companies could just hop jurisdictions because the state turned against them and their customers is an insane one.

Re: Identity verification on Claude

#663

Post by sofixa »

Aurornis wrote: post #438
dalemhurley wrote: post #432

Earlier quoted context omitted.

Why, is there any evidence to support your claim?

There is a long history of US export restrictions on technologies. They have all been temporary. It would be extraordinary if this was the beginning of the first ever permanent technology export restriction.

It doesn't matter, because people are adopting AI now, and even if the restrictions are lifted tomorrow, it has been proven that relying on an American SaaS is risky, because you can just lose access.

So why would you get that instead of self-hosting Mistral Studio and fine tuning for your needs? Or going fully DIY with e.g. OpenShift AI and open source Chinese models?

That way you fully control availability, costs and quality.

Re: Identity verification on Claude

#665

Post by GL26 »

If models start only being available for a certain class of citizens of x,y,z countries, there will be "smart hackers" selling out their token usage for twice the price or something (a litteral black market for tokens).

Re: Identity verification on Claude

#666

Post by sofixa »

palmotea wrote: post #622
sofixa wrote: post #616

Earlier quoted context omitted.

No no, the remarks were pretty damn fucking serious. And the fact that the US could elect such an unserious and demented criminal, twice is proof enough that there is a lot to worry about. Who's to say who will be next? Or after that?

> No no, the remarks were pretty damn fucking serious. Trump says a lot of things. It's foolish to take every outrage seriously. It's pretty clear it's just a tactic of his (albeit a dumb one). > And the fact that the US could elect such an unserious and demented criminal, twice is proof enough that there is a lot to worry about. Who's to say who will be next? Or after that? Ok, then. If you're serious: kick the US o…

> Trump says a lot of things. It's foolish to take every outrage seriously. It's pretty clear it's just a tactic of his (albeit a dumb one).

And he has also acted on a lot of the things he has said. It's foolish to ignore him.

> Ok, then. If you're serious: kick the US out of NATO. Increase military spending, go it alone. Really treat the US as a non-ally.

What? It's not possible to expell a country from NATO as per its charter. And European countries are increasing military spending, by and large buying, in that order, domestic, European, major NATO aligned suppliers (most notably South Korea), and the US only when there is no choice (so F-35 and munitions/replacements for existing systems). And of course this ignores that current European military capabilities are plenty for European strategic autonomy.

> Or, you know. Wait for him to leave office or die (he's 80!), and write the Democrats and tell them to get their fucking house in order. Trump didn't win so much as the Democrats were utterly incompetent.

I don't think you're getting it. Fool me once, shame on you. Fool me twice, shame on me. The trust with the US is irreversibly broken and not getting repaired in the decades to come. The American political system is broken beyond repair, and even the opposition to the current wannabe fascist regime are not campaigning to change anything about it. This means that regardless of who's in power next, even if they're the best thing since sliced bread, there is no guarantee the following elections won't bring another bout of wannabe fascists threatening to invade their allies.

There is no going back from that without major political reform in the US and years to earn back trust. AKA not happening. Even the opposition, even the most radical opposition, venerate American political institutions as the works of deities that can only be slightly changed, not replaced wholesale. And, from the outside, the whole system - how elections are done (you need proportional representation, term limits, no janky districts), representative structures (flat two seat per state senate doesn't work anymore, house of representatives having skewed and fixed numbers doesn't either), judicial system (political appointments and elections, and for life appointments, do not work), supreme court (political appointments and for life appointments do not work), etc are overdue by decades for serious structural reforms.

Re: Identity verification on Claude

#667

Post by xinayder »

ozozozd wrote: post #576
Aurornis wrote: post #437

Earlier quoted context omitted.

Despite what’s being implied everywhere, this ID check page has been there since April. Wayback Machine if you don’t believe me: https://web.archive.org/web/20260415064244/https://support.c... > As a non-US citizen Opus 4.8 is the best American LLM I will ever have access to. That's no longer up for debate or question. This is a crazy conclusion for a situation that isn’t even two weeks old. LLMs are not the first te…

This is a strange plea for optimism. Sincerely or not, judgement is yours, Dario has been begging for regulation. He has been talking about how Claude models are distilled by foreign adversaries. And now the regulation is here. What makes you think this situation that the CEO of Anthropic is asking for it temporary? Do you not believe Dario was sincere?

How ironic it is that multibillion AI companies are complaining that their models are being distilled (read: used without permission) while the current top players trained their models on stolen data...

Re: Identity verification on Claude

#668

Post by fc417fc802 »

gentooflux wrote: post #650
fc417fc802 wrote: post #483

Earlier quoted context omitted.

That's nonsense. You might not be able to afford to enter the market but that's not at all the same as a physical constraint capping the number of competitors.

You might not be able to afford to put up power lines or build a road, but that doesn't mean you physically cannot. By your logic there are no natural monopolies

If I want to service 1234 sesame st from my physical location with power lines there are a very limited number of routes I can take. The limited physical space quickly becomes crowded, duplicate physical infrastructure of that nature is highly inefficient, and there's a (completely pointless) negative impact on society to accommodate the constant installation and servicing.

Commodity hardware installed in a commodity building on commodity land using commodity power hooked up to a commodity network isn't the same thing. By your logic anything expensive or inconvenient is a natural monopoly. I believe the term you were looking for was "high barrier to entry".

Re: Identity verification on Claude

#669

Post by yurishimo »

joxdosba wrote: post #644
interloxia wrote: post #627

Earlier quoted context omitted.

Where possible, I believe it's our duty to educate those who know or care little about how their devices work. Unfortunately stats aren't persuasive for my family, but they might be for some as a starting point. https://bjs.ojp.gov/library/publications/victims-identity-th... Victims of Identity Theft, 2021 October 2023, NCJ 306474 https://bjs.ojp.gov/document/vit21.pdf

These links claim that identities are stolen, but do not explain how. What happens to the victims, who are now presumably left unidentifiable? How are they tracked if they can’t be identified? Do their families recognise them? How does that work if they were married, had children or something? Does the identity thief just take over their whole life? As far as I know, “identity theft” is a boogeyman invented by the ba…

I would maybe reframe it as fraud being a natural downstream effect of identity theft. I don't need to steal anyone's identity to forge a check but that's also categorized as fraud.

Identity theft also paints a clearer picture of what is required to remedy the situation. If your details have been pwnd hard enough, you might need to get new government documents entirely in order to protect yourself long term.

If a contractor takes my money to install a pool and then disappears, I don't need to reset my entire financial identity. I think it's worth having a separate idea to describe the situation as a whole and not just the specific vector in which a crime might have been committed.

Re: Identity verification on Claude

#670

Post by teliosix »

ElProlactin wrote: post #571
hoytschermerhrn wrote: post #558

Earlier quoted context omitted.

I’m a US citizen and absolutely will not be uploading additional information just to use a company’s models. This effectively kills my usage of anthropic for anything beyond their 4.8 models.

> I’m a US citizen and absolutely will not be uploading additional information just to use a company’s models. I can't speak to your specific use of Anthropic's models, but I find it interesting that people will identify themselves (to set up and pay for an account) and provide all sorts of personal (and often sensitive if not confidential) information to these models on a daily basis, but balk at a 5-minute identity…

It is just the standard performative outrage over and over. The same person will be uploading their ID a few days after it is introduced.

So many people have internalized an anonymous audience that they are performing to in terms of what they think the anonymous audience wants to hear that they can't tell what is their own actual thoughts and motivations.

That is why the motivation changes two weeks later and the ID gets uploaded.

Post reply on HN