1. Release fable, highly nerfed and limited
2. See the compute capacity limiter pegged day after day
3. Lobby to the government, claiming ai is super unsafe and not aligned and they must do something
4. Government "forces" anth to turn off
5. Anth takes the pressure off of compute capacity, and gets to blame it on the govt
Like you're telling me fable is somehow an order of magnitude better than GPT 5.5 to the point where it compromises national security, despite evals and anecdotes saying otherwise? Nah.
You're saying a company's marketing department can casually get the United States Government to issue a national security passage, preventing sale or distribution of their product?
Was their ongoing designation as a "supply chain risk", which they are suing to overturn in court, also a marketing stunt?
Seems like a really strange thing to use that sort of power for - why not just get all your competition declared persona non-grata and seize monopoly power?
Oh, cool, then surely you can point me towards the posts where they're celebrating this, or even actively advocating "please ban our product on a Friday with no notice or due process"?
In interviews, this week alone, Dario has gone on record repeatedly about wanting to slow AI progress.
Anthropic silently degraded AI-research queries to Fable (they changed course on this, but they still thought it was a good idea)
And now that the government is taking them at their word, they're trying to drag GPT-5.5 and OpenAI down with them
Yes, the administration is being heavy-handed, but unfortunatley it's the logical end of telling everyone you built a "nuke" and that it's possible for people to use it against us
Oh cool, we're back to "she was asking for it, wearing a short skirt like that".
There's a difference between regulation and the government randomly waving it's dick. The government was aware of Mythos capabilities since it was announced two months ago. If there was ANY sort of serious concern, they could have just blocked the release. Instead, it got pulled three days after release.
We would be having a very different conversation if this was an actual regulation, which is very clearly what Anthropic was actually asking for
Now leave the nice lady in the skirt alone, you pervert.
Yes, it's marketing straight out of the Big Tobacco playbook. Convince everyone your product is dangerous, get government to ban it, and then... uh... pivot into adjacent market segments?
Not a good analogy. Fable/Mythos are unsustainably expensive for Anthropic. They want/need the world to be nerfed to remain solvent. They didn't have to turn off -all- models
Incidents like this show how unenforceable GDPR is, and how it's been a net negative for users since its inception. It's idealogical back-patting, toothless when it matters.
After the GDPR every website added an option to export your personal data and to delete your account. Something most were missing at the time. It was an immediate and massive win.
Right, but nothing stops companies from refusing SARs on baloney grounds.
Complain to a DPA? They tell you to go through ADR or outright ignore you. Complain to Ombudsman? They'll tell you the same. (In my experience, the Dutch do this)
Company ignores ADR? Sure, now you can go through the legal route and spend copious amounts of money all because a multi billion dollar company knows the game and how to navigate the bureaucratic mess better than you.
Yep, this is how they do it. The domain registrar netcup did something like this to me.I went through their parent company (?) too, without success. They will put forth any reason to not have to delete your data. I suspect, that they either are trying to reduce work for themselves, or their platform is so crap internally, that they would have to get someone coding to delete the data.
This. In reality, GDPR isn't preventative, nor punitive enough for any meaningful user protection. We get cookie banners everywhere and user data harvesting companies happily pay the negligible fines
The DPC would disagree. All you need to show is that you took "reasonable steps to protect users," which is trivial to do, and not even a single fine will be levied.
One of the most frustrating parts about Google's approach to AI in general is their project manager-y directive from on-high, that any Google product needs to adopt all Google's AI tools, wherever possible, and will be ranked on how deeply the integration goes.
> and will be ranked on how deeply the integration goes.
Oh, that must be why the Gmail web interface has so many spans that contain nothing but another span (including the span containing the "Press / to help me write" placeholder text in the message body area when replying). Here you go, boss, it's deeper now!
Codex's computer use came from OpenAI's acquisition of the Apple Shortcuts team, whose institutional knowledge allowed them to exploit all sorts of undocumented macOS APIs, not some virtuous accessibility* stack. With 99% of work happening on the web anyway, it IS fair to say that it's not the year of the Linux desktop, or any desktop, because the desktop doesn't need to exist at all.
*macos26 introduced a multitude of accessibility regressions that have real-world impact on humans with disabilities, let alone AI
It doesn't mind what Chromebooks have under the hood. They don't make the Year of the Linux Desktop closer rather the another Decade of Chrome browser.
It seems like the issue wasn't the pressure so much, but the temperature. The contents were supposed to be stored at 50 degrees. The thermometer for the tank was reading 100 degrees, but nobody knew what the actual temperature was, because the thermometer's maximum reading was 100. So the challenge was to figure out how to cool things down (which would relieve the pressure in turn)
Any change in the past 12 months is abolutely not up to the Cayenne, which came to market 24 years ago. Or the Macan, which came to market twelve years ago. And that was the main point of the GP post--how adding SUVs to the mix didn't ruin the brand.
But in any event, Porsche sold more cars in 2025 in North America, than any year prior.
It did take a big hit to profitability in 2025, mostly on one-time restructuring charges. But until then it was one of the most profitable auto-manufacturers out there. And its annual revenue is still higher than most of its history, especially on a per-car basis.
So sure, recently Porsche hasn't done well. But it has very little to do with SUVs and that transition. And I would argue that the brand itself is still very strong, even if operationally they have mishandled the electric transition.
> But in any event, Porsche sold more cars in 2025 in North America, than any year prior.
Porsche sells about as many cars every year as Ferarri has sold in its entire existence. I'm not sure that's a strong indicator of whether or not it "brand" (AKA public perception) is doing well or not. Clearly Ferarri has a strong brand than Porsche, despite only selling 330,000 cars in the past 80 years. And despite Porsche selling 310,000 in 2024 alone.
Yes they have very different business models. But it would be like using "number of Window's licenses sold" to argue that Microsoft has a really strong brand right now.
Which is why I was comparing Porsche's 2025 sales data with its own prior years' sales data and not with Ferrari's sales data. Year over year is an imperfect but reasonable-directional proxy for brand staying power, especially given that the product mix didn't change all that much (except that the 718 series were put on hold, which one would expect would make it worse, not better).
"While loyalty has fallen slightly since last year’s study, some brands held strong with buyers. Porsche was the top premium car brand with a 58.2 percent loyalty rate, followed by Mercedes-Benz with a 49.7 percent rate. Lexus ranked highest in the premium SUV segment, with BMW a close second."
For Microsoft, number of licenses sold year-over-year would be relevant. Although the comparison isn’t great because it is harder to switch operating systems than car brands.
I'm working on a parenting tech device and the enclosure for it is completely AI generated. I hadn't a clue where to even start with 3D modeling, and an LLM taught me that it's code like anything else.
Weirdly, Opus 4.5 one-shotted it perfectly, but this was right before the nerfing controversy, and it's been very difficult to make even minor tweaks to the enclosure ever since.
It's like Opus went from an expert shape rotator to not having any idea what it's working on.
1. Release fable, highly nerfed and limited 2. See the compute capacity limiter pegged day after day 3. Lobby to the government, claiming ai is super unsafe and not aligned and they must do something 4. Government "forces" anth to turn off 5. Anth takes the pressure off of compute capacity, and gets to blame it on the govt
Like you're telling me fable is somehow an order of magnitude better than GPT 5.5 to the point where it compromises national security, despite evals and anecdotes saying otherwise? Nah.
reply