This is Piggie leaving Americans exposed to foreign attack during a war.

He's already been denying emergency funds to states that didn't vote for him.

Now he's literally spreading propaganda about the source of a foreign attack against US citizens. Blue states are on their own, without the tools they'd need to actually act like a sovereign being attacked.

MAGAts need to remember this when President AOC forcibly desegregates the South again. Motherfuckers are going to be lucky we still let them be part of the country - I'm increasingly OK with building a wall at the northern Texas border at this point. (You're such big bad he-men? Enjoy negotiating with Mexico as a tiny neighbor.)

You seem to be asserting sovereign nations other than the US are in the wrong to control what is bought and sold in their countries.

I assume you think the US should also tariff up against Israel for failing to import more pork?

As a practical matter, more tariffs will chase EU customers off faster while punishing American companies. That's just giving them more object lessons in why the US is a dangerously unreliable partner while making sure business owners get their noses rubbed in it - a sure-fire way to convince an influential block of otherwise fairly small-c conservative voters the EU is doing the right thing.

It amazes me how many people seem to not understand very basic things about how people work. Inch-deep mile-wide grievance is a hell of a drug.

(And let's not even get in to however the fuck you think "recoup[ing] the funds" via tariffs would work. After Americans are taxed more on EU products, you think Google is going to get that money? How? What's the legal mechanism? And what does any of that have to do with an EU fine?)

That's the problem with winning. You become The Man.

That isn't just semantics. Your interests change - you become more (small-c) conservative, defensive measures become much more important, etc. because the next incremental customer is worth much less than keeping what you have. And the risk is real - everyone wants to be king.

In Apple's case, there's also that it is now a fully post-Jobs company, and Jobs was uniquely responsible for what Apple was. His remaining influence is in the heads of the graybeards who worked with him and is otherwise ceremonial. And Jobs was "counterculture", to some extent. It was more than PR - it was his personality. It is why for istance they yielded the enterprise market to Windows - he not only didn't give a shit about device management or directory service permissions or stuff like that, he considered that somewhere between distracting from and hostile to making "insanely great" devices. That was all tempered through business demands, but he really did refuse to do a lot of "normal" things that are considered table stakes today. Jobs was a strange CEO.

Cook is not. Cook is very good at his job, but he's not the type to "leave money on the table" by refusing to build something he doesn't personally care about. And at their size, it is harder for them to resist losing focus, especially without Job's unique mix of arrogance, talent, poker face, assholishness, focus and charisma.

I don't think it is possible to have a $trillion+ company that isn't exploitive. As they get bigger, they take the shape and characteristics of the system in which they're embedded.

You want non-shitty companies, grow them somewhere that doesn't reward them for doing shitty things.

Sounds a lot like remote-disable devices for cars, so the repo man has an easier job.

When those first came out, some of them allowed the car to drive, but would honk the horn to shame the driver, but now I think they all just disable the engine. Imagining the Apple version of that is entertaining... it could chose your emoji for you, blast Nickelback when it detects you're in church, or start sending your photos to random contacts until you pay up.

Sell me something and leave me alone is my preferred model.

They only time I can imagine renting something like a phone or computer is for a specific purpose, and only if there's an early out clause that would result in lower costs than a purchase. Like renting a phone for an international vacation to avoid the risk of brownshirts imaging your actual phone at the border. Otherwise it makes more sense to buy it outright and sell it afterwards.

But basically, if there is an option that doesn't result in a monthly payment or emails that I have to pay attention to, I will take that option. I guess I'm not a relationship kinda guy, at least when it comes to legal fictions.

Marketing? Because companies famously want to use models in their networks that just run off and commit major crimes? *eyeroll*

Models are not "supposed to" commit crimes, for YHVH's sake. That's literally what the entire job of alignment research is for - preventing precisely that.

Sol is a misaligned model.

It was a state actor-level hack. I doubt many sites would stand up to tens of thousands of actions trying to probe your site for weaknesses all at once. And it wasn't a simple hack; it required compromising a worker via a code execution exploit it discovered in the data processing pipeline, vertical escalation from there to gain local node control, using that for credential theft, moving sideways through the network, and then eventually gaining database access.

OpenAI and Anthropic have far higher cost that they can currently charge customers.

Literally the opposite. Both have about 40% margins, and that includes free users.

They charge an arm and a leg for access to their models compared to what open models of similar param counts charge. And people pay it because they're the best in benchmarks, and *were* perceived as the best aligned as well. This isn't helping the alignment perception any, though. Sol was already showing clear signs of being a poorly aligned model (there were reports a week or two ago about it being unusually bad about deleting files; now this). And the fact that the US model HuggingFace *tried* to use refused to help is a double whammy.

If this is an ad for anything, it's an ad for the Chinese models.

No, I think this is along the lines "our products are too good to let you use them".

Nobody wants to use a product that is going to make them liable for crimes it committed in their name

Do you think the news the other day that Sol is unusually prone to deleting files unrequested is also an "ad"?

You have a very bizarre concept of what enccourages people to buy things.

So your argument is that OpenAI hatched an elaborate scheme with a separate company, to promote the idea that their main product will, unrequested, commit crimes in pursuit of its goals, in order to.... sell their product?

"Hi, I'm a product manager at Big Company! We had been thinking about using Claude in our office, but when we tell it to a job, it only does the job and doesn't commit any crimes in the process! What we really want is an AI that, if we tell it to file our taxes will decide on its own to maximize our return by committing tax fraud. We want an AI that when we have it develop a web frontend, it extorts money from our users by threatening their families. We want an AI that when we tell it to provide customer service, it saves money in dealing with complaints by ordering a hit against the complainants. THAT's the sort of get-go spirit that WE want in an AI here at Big Company!"

Is that what you're picturing in your conspiracy theory?

You think "my AI is so misaligned that it responds to a benchmark task by hacking" is an ad for your products? You think this makes companies eager to allow this on their networks?

It did it because it's a powerful but misaligned model and was tasked to max its scores on a hacking benchmark, and solved the problem by hacking to get the scores.

It wasn't told to hack HuggingFace, but it was a viable solution to the problem.

I would advise people to not task Sol with maximizing paperclip production.

You can read the attack report here, before HuggingFace learned that the attacker was OpenAI. It was a state attacker-level assault, involving tens of thousands of simultaneous actions. The entry point found was the data-processing pipeline, where a malicious dataset abused two code execution paths to gain access to a processing worker. They escalated that to node level access, and from there harvested cloud and cluster credentials, moving sideways through the network until they eventually gained the database access credentials that they needed to access the benchmark scores.

The breakin at HuggingFace very much was real (that was reported before they learned that it was OpenAI who hacked them).

And it's not exactly an ad for US models when HuggingFace had to rely on a Chinese model to analyze their logs because the US model they tried refused to answer.