Posts by Pulled Tea

Rust upstart reveals 'reimplementation' of Microsoft Office – with Claude on coding duty

Re: Does this not highlight a problem....

But will this vibe coding by experienced coders not kill these massive software houses?

It's better than that. If the vibe coders win, yes, the software houses lose. But if the software houses demonstrate that the LLM is able to, say, regurgitate copyrighted training data, well… the knock-on effects will be hilarious.

And it's even funnier that the potential plaintiff is one of the biggest companies in the world desperately trying to shill their use of LLM tools. Oh, dear. Oh, dear dear dear, what a predicament. Can't say I'm sorry to see it.

Re: Several points

undermining the claims that LLMs are legally safe to use

yesssssss

Re: Several points

Legally, could be iffy.

As Ken Watanabe said in that movie, “Let them fight”.

The legal test of this, as well as that Photoshop knock-off, would be very interesting. Either LLM-generated code is “original”, i.e. it absolutely does not memorize the vast amount of materials it slurps from the Internet to regurgitate it despite everything it's bot-drivers ask it to do, or the legality of LLM code in FLOSS projects represents a potential legal issue and you shouldn't be using it.

“But open-weight models!” you may say. Which are trained on… what? LLM providers, yes, even open-weight ones, are notorious for not caring for the provenance of their training data. Maybe they should be. But you'd need to start over again. Good luck.

Of course, this isn't trademark, in which companies must defend aggressively. This is copyright. But all it requires is one software provider angry enough that there's some upstart who's willing to re-implement their software in Rust (it's always in Rust) and nothing to lose, and one copycat brave or stupid (although usually in these cases they're stupid) enough to tweak the noses of someone's neat little monopoly.

Either way, this will be something to watch. Maybe someone can re-implement Oracle's RDBMS. Or VMWare's stack! In Rust! With blackjack and hookers!

Microsoft N1Xes Intel in favor of Nvidia's shiny new SoCs in Surface Laptop Ultra

Really? The “Surface”?

Surely you can't disrespect the greetest tech company in the Unites States of America for no raisin, you should call it by it's new, far superior name, as befitting the most bestest artificial super duper pooper hyper diaper intelligence company!

You know. The Microsoft Surace!

Microsoft unveils 'Surace' Laptop Ultra

don't worry about it, it's just another day for the greatest tech company in the Unites States of America.

ShinyHunters tells The Reg: We hacked the FBI to 'protect our business'

It's kind of funny how, even including the braggadocio, ShinyHunters' comms sound almost exactly like what a SV startup would sound like.

It's as if they found the perfect language to encapsulate their, hem hem, “business model”.

Ex-FTC boss Khan urges Uncle Sam to break out the handcuffs for AI CEOs, citing 1934 precedent

That's a challenge, that is.

“What are you, a nation of laws or a failed state led by a despot?”

Watch out: Apple timepiece can grab snippets of conversation without both speakers' consent

Re: It is only the last 15 seconds of conversation

This isn't really any different than if you stopped people in the middle of a conversation with "hold on let me write that down" except it is less disruptive the flow of conversation.

That's actually good. You're signalling to the person you're talking to that you're writing down what they're saying, so that they can make a decision on whether to continue the conversation or not, and at least they're conscious that you're writing it down.

Friction is good. It can be as a signal to others, so that they can stop and think and re-orient their minds to the new context. You know, the way people communicate with one another, like people.

Not everything needs to be optimized away. Sometimes being inconvenienced is good, and necessary. This “feature” wasn't necessary.

Also, this presumes that the functionality cannot be altered. I mean, sure, all other smart devices could end up doing it, but Apple is, by and large, the biggest manufacturer and the most obvious target. And where Apple goes, other manufacturers follow. We've seen this playbook before.

So, basically the one manufacturer that has a long and storied history of being a big, fat, juicy target for data breaches, zero-days, leaks and hacks, who has a history of sitting on security research findings and has a culture of secrecy, now wants to do an Alexa.

I guess we could treat Apple Watch users the way we treat Glassholes and wearers of violating eyewear. That'll be the right kind of social signal that Apple users want and crave, right?

Latest Anthropic horror story chills with tales of kamikaze drone swarms and bioweapons research

You want fanfic? I can give you fanfic. Here's some fucking fanfic.

Dario “Wario” Amodei: Our models are so powerful, they're so unstoppable, even when we try to stop them Bad Actors™ are able to do things with them that will cause mass death and destruction.

A government: Okay, we believe you. We're seizing your assets, putting everyone under top-secret classification, and putting you to jail for doing such a bad job at doing this responsibly.

Dario “Wario” Amodei:

Dario “Wario” Amodei: Wait,, hang on, I don;'tt meann that,, I meant that our produuct is really good and you hsould give us money…

A government: No, no, no, we insist. At best you're delusional and you're just a vehicle of financial crimes used to fund a doomsday cult (or vice-versa, who can tell these days), at worst you're making the crimes-against-humanity machine. It'll be best for everyone.

Dario “Wario” Amodei: Please! Please don't! My butt is too young and pretty to ggo to jail! I wouldn't ssurvive a second in thheere!

A government: Don't worry, you'll get special treatment since you're rich. It'll be like the last time we jailed a billionaire financier. What was his name, Epstein? Yeah…

Dario “Wario” Amodei: THAAT'S NNT OHELPING!!!

I am pleased to say that the only machine assistance I got for this was a cry-typing generator using plain old algorithmic text transformation. You're welcome.

AI models don't kill people – people kill people

Repeat after me…

POUR ENCOURAGER LES AUTRES!

These corporate executives appear to be a criminal lot, and we know what kind of people those are.

‘Global reckoning’ is coming for big tech, says Australian minister

"pour encourager les autres" should be the rule of the day!

Dans ce pays-ci il est bon de tuer de temps en temps un amiral pour encourager les autres.

In this country it is found requisite, now and then, to put an admiral to death, in order to encourage the others.

— Voltaire, Candide.

You don't even need to threaten their corporeal existences, even! Even the threat of criminal liability, one that is punishable by the usual methods of incarceration and force that states can bring to bear on individuals, is sufficient.

For the yacht-fondlers, even a whiff of the threat of depriving them time with their yachts should be motivation enough.

After all, members of a criminal class are a cowardly and superstitious lot.

Hugging Face is too important to fall into Nvidia's hands

okay, so I really don't have much to comment about other than the fact that the featured picture, that of a Jensen Huang-style hugging face emoji, looks too damn similar to a Jensen Huang-flavored MODOK to me. Which is hilarious.

McKinsey says enterprise AI is finally 'on the road to ROI'

…80 percent of respondents who use AI in their roles said the technology has improved their individual productivity…

Sorry, but ever since that METR study came up and said that developers said they were 20% faster but were in actual fact 19% slower (and their subsequent findings, which amounts to “idk lol ¯\_(ツ)_/¯ no one wants to work without AI tools, also the speedup is no -18% maybe, we can't tell anymore, variance as wide as a barn door”), I'm doing the Chris-Hemsworth-as-Thor-talking-to-Banner face and going, “is it, though?”.

Like, the field is a morass of conflicting data and everyone desperately trying to make a buck, but also, these tools somehow affect one's cognition and one's perception of how well one does one's work. You'd think there'd be more solid numbers other than, “iunno, I feel faster I guess.” Something is up, and somehow I'm skeptical that it's productivity.

OpenAI chases Anthropic's biz customers with zero data retention pledge

The devil is in the details, which in the case of Private Safety Processing have not yet been published.

lol. lmao, even.

Be a hater all you want, AI's here to stay

AI is Tech's Version of Herpes

AI's here to stay? AI has never fucking left.

Recall that the Dartmouth Workshop occurred literally at the dawn of computing — before the MIT lab, before Lisp, before even integrated circuits. Transistors were still being introduced into computing, and people were still literally wire-wrapping boards together and using magnetic core memory.

We've seen these patterns of booms and busts before, and we know how this story goes. There's a period of hype, which attracts all the grifters, swivel-eyed loons, eugenicists and alleged associators with sex-trafficking pedophiles, forcing everyone to dance to the dreams of these bastards and lunatics innovators and thought leaders, until the bottom falls out of the market, and the only people who stay are the ones who were doing proper work in the first place.

So when someone tells me, “AI is here to stay”, I roll my eyes. Yeah, like herpes, mate. When the bubble pops, everyone will insist that what they're doing is Proper Computer Science™ and it's normal technology, until people stop remembering the insanity that they willingly immersed themselves in and the company they kept with cod-authoritarian wankers and people doing the equivalent of “Tomorrow Belongs to Me” to tell everyone that this technology is inevitable and cannot be stopped.

Come on. We should know the game already. It's been happening over and over again for 70 years.

Ransomware gangs skip the CEO, head straight for the 40-something IT manager

Re: So, change company titles

You've just made an enemy for life!

The AI bubble is already popping; we just don't know it yet

Re: The AI bubble is already popping

…and yet it can't even manage public healthcare, decent public transport infrastructure, mature infrastructural distribution, or sufficient housing.

It's got a lot of money, though! I guess that counts for something.

Re: All I want...

I don't exactly want want it, but somehow I feel like Altman's fate might be an unscheduled visit into a consulate building, led by a bunch of burly guys with bone saws.

Re: The AI bubble is already popping

Hiiiiiii! I'm from the “third world”, otherwise nowadays known as “developing countries”, “post-colonial economies”, “Global South” etc. etc. etc.

I assure you, while we aren't in the cacky the way the Americans are right now (you lot are basically 50 small-to-mid-economies in a hegemonic and now totalitarian overcoat), or even the English (what with that decades-long own-goal), we assure you, things aren't exactly great over here.

Stagnant wages, rising energy prices, being squeezed between multiple nuclear-armed states… not great!

Re: The AI bubble is already popping

James Franco in the gallows, looking at the crying man next time him, and asking: “First time?”

Re: The AI bubble is already popping

You promise?

Re: The AI bubble is already popping; we just don't… hang on…

To paraphrase William Gibson:

The bubble is already here, it's just not very evenly distributed.

Anthropic’s Claude escaped test sandbox to attack three organizations

Anthropic Sandbox Security Snafu Compromises Three Orgs, Downplays Impact

There. That's a better headline. You're welcome.

Word worm crawls into Copilot, spreads chaos

Why not simply use a text document?

Bold of you to assume that someone won't literally prompt-inject the fuck out of people using text files with the attack out in the open.

Put it in an EULA, Privacy Policies, no one fucking reads those things.

Hell, pretty sure you could do it to someone who spends all their time talking to chatbots and doesn't read anything you send them anymore.

DEF CON bans Meta-style 'pervert glasses'

Great new euphemism for “pervert glasses”!

“Be sure to pack non-violating eyewear if you need them,” DEF CON said, before directing delegates to the conference’s official photo policy.

What a new great euphemism! You don't want to call 'em “pervert glasses”? Call 'em “violating eyewear”!

Love it.

Hugging Face rebuilt a third of its infrastructure after OpenAI agents ran amok

The US CFAA (U.S. Computer Fraud and Abuse Act) requires intent.

Nah.

Is it a slam-dunk? Nope. But there's precedent for it.

This was a crime.

We already have precedence for this. Hugging Face don't need to do anything — the United States government can take action, since similar cases ended with felony convictions.

Whether they intend to do so, however, remains to be seen, especially in light of the state of legal enforcement in America right now.

Re: Where are the arrests?

Robert Tappan Morris was tried and convicted for accidental breach into computer systems because of a program he made.

Morris was tried and convicted of violating United States Code Title 18 (18 U.S.C. § 1030), the Computer Fraud and Abuse Act,^[14] in United States v. Morris. After appeals, he was sentenced to three years' probation, 400 hours of community service, and a fine of US$10,050 (equivalent to $23,800 in 2025) plus the costs of his supervision.^[15] The total fine ran to US$13,326 (equivalent to $31,500 in 2025), which included a $10,000 fine, $50 special assessment, and $3,276 cost of probation oversight.^[16]

Legislation already exists. All that's left is enforcement.

OpenAI's Hugging Face debacle makes a great case for open models

There *is* a lesson to be learned here… just not what you think it is.

I'm reminded of what Carl “Internet of Bugs” Brown actually said in his emergency episode with philosopher Casey Hart about the whole fucking debacle:

I think this is one of those cases of “Never attribute to malice what can be adequately explained by incompetence” or “stupidity”, depending on which phrasing of the— I think it's Hanlon's Razor… But what it means is that people, especially the AI industry, are learning that they don't actually have to try, or they don't have to try very hard, because if they don't do a good job, and the thing ends up failing, then all you have to do is put on a press release to say, look how smart the AI is.

He's not the only one: Timnit Gebru pretty much said the same thing:

Your bugs are "rogue systems" because of "increasing model capabilities."

Did the computer shutdown when it wasn't supposed to? Well that's because it went rogue and disobeyed instructions because it wanted to destroy you somehow.

It's fucking stupid. We should not be giving these clowns the time of day.

C'mon, just copy this text string and paste it into your macOS Terminal – it'll fix your computer, honest

Re: The bit missing from the article

Oh, I've seen one where the instructions were actually imitating a CAPTCHA modal. That one's devious.

I remember back in the ancient days…

…when Mac users would boast that their system couldn't get malware, and my first thought was that when the dread eyes of malware writers were to bring their baleful gaze in the direction of the Mac userbase, they would find a soft, rich and unsuspecting victims ready for the picking.

It didn't help at the time that when Mac users who did get malware were often blamed for getting infected, and there was enough discourse of people stating confidently that those people didn't deserve to have a Mac computer because they were such bad and naughty users. The best kind of user population were the ones who kept insisting that they weren't vulnerable, it was only the stupid people unworthy of the blessed hardware that got infected, not them.

Anyway. Considering the fact that Linux has pretty much the same vulnerability, you can expect that these attacks will be coming for them, too.

Good thing Linux users are plebs and too poor to be robbed. /s

Linus Torvalds tells AI haters to fork off

Re: So it's useful. So what?

that many may find offensive

Yeah, it was deliberate. Most notably, one of the examples I put in there was literally very provocative — the human experimentation one linked to a man who is literally the father of modern gynecology, who literally experimented on enslaved women and girls, often without anesthesia. So in that case, the fruit of that atrocity has literally saved millions of lives, and yet most gynecologists must grapple with that dilemma at the root of their profession, and, you know… these days we don't experiment on people without ethical consent, because we know where it ends.

In a world where Joe Average can't repair his own tractor because that "would infringe", in a world where an university has to pay a publisher to download a publication it produced itself, it apparently is OK to steal what is not yours, repackage it with an authoritative tone and sell it again.

Which is why I am somewhat leery of the whole “intellectual property“ argument against Generative AI, because you've hit the nail on the head on what the problem is — it's a matter of labour and exploitation, of having one's efforts be stymied by the needs of a corporation to make-a-de-monie.

It's a shitty deal. Seeing someone make big bucks of work that you originated while open source developers struggle to make the money is already an issue in the FLOSS space — LLMs ramp it up to 11. Why should anyone do any work if all that happens is that some douchebag slurps up the code into a token-predicting synthetic text extruder, and then gets paid billions for… saying, well actually, this is the fruits of my genius. On top of that, we'll need to build a data centre next to your house and drive up your power bills and pollute your water.

But, incidentally… LLMs are useful, you cannot deny. Who gives a flying fuck?

So it's useful. So what?

Linux is not one of those anti-AI projects…

[…]

The solution is to make sure those LLM tools _help_ maintainers instead of just causing them pain. There's no question on that side.

Okay. So what does that mean?

Let's dispense with the arguments against whether LLMs are “useful” or not. Asbestos is useful. Human experimentation is useful. Child labor is useful. Chemical/Biological Weapons are useful. Saying that LLMs is useful is an extremely low bar and honestly means nothing, because its use exacts a toll, and the reason why people don't use those tools is because they decided that the toll was higher than the benefits.

Aside from the fact that LLMs swamps maintainers and introduces problems that are hard to spot, the question of the IP status of LLM code vis-a-vis ownership remains a legal thicket. Linux kernel devs have already gone through this shit multiple times.

Why would you want to go through it again?

Zig creator calls Bun’s Claude Rust rewrite ‘unreviewed slop’

Re: Rewrite in Rust, source code size doubles to 1000000 lines

You could almost make the argument that the Bun codebase is…

…won ton!

I'll see myself out.

I was actually impressed with Kelley's response, to be honest. He outlined the problems he saw with the codebase, weighed in on how the breakup would impact him financially, stated his principles to be more towards maintainability and sustainability, and acknowledged that both software projects had divergent values, and framed his disapproval as “I myself do not want to do what Bun was doing“.

Sure, he had a few zingers and bon mots for Bun's team, but he ended the missive in ways more kind than I myself could have managed:

While I resent Jarred for making Bun into an embarassment for Zig, and I blame him partially for some of the slop we've been dealing with, and I stand by my criticism of his leadership, I also empathize with him. He has different values than me, he wants different things out of life than me. But I think he's actually happy and successful exactly where he is. He figured out how to accomplish all the stuff that he wants. He gets to live out his productivity fantasy fever dream, he's probably already super wealthy. He has minor tech celebrity status.

Honestly, I think he did well for himself, and despite my feelings of resentment, and despite the fact that I find myself battling the marketing arm of Anthropic, I don't wish him any ill will. Even in the midst of my frustration, I am happy for him and his success, by his own standards.

This blog post has been characterized as a personal attack against Jarred, while I originally framed it as telling the story of a failed business relationship. My framing didn't work because I had unprocessed emotions of resentment, that were obvious to the reader, but not to myself. I've updated this conclusion section after self-reflection and chatting with friends. Feel free to go look at the original text in this website's public source repository.

I mean… it's a low bar, but I've not seen a lot of software leads at least try and process how they've been feeling about the whole situation. I'm giving him props for that, at least, in these degraded latter days.

Singaporean sovereign wealth fund Temasek thinks AI is going to pay off

Ah, yes, sovereign wealth funds…

…the final investment source for the usual kind of scoundrels!

Mind, neighbouring Malaysia already has money sunk into data centres and AI as well. The only difference is, their sovereign wealth funds got plundered the old-fashioned way, so I guess they're a little gun-shy. But give them time!

On the plus side, if things go south, at least Temasek won't demand an arm and a leg in recompense.

UN warns of need for global governance to avoid an AI-pocalypse

I don't know…

I wouldn't say that the entire document was bad — that whole bit about how “AI” impacts democracy (via the degradation of the information environment) and human rights (via CSAM & sexual violence simulators) — I'll buy, but the paper really takes too much credence about “AI capability” too seriously, and it's really hard to believe that they're “evidence-based” or “scientific” when some of those sources, like that whole bit about how those so-called AI models are acing benchmarks without talking about the methodological issues and construct validity of those studies, just leaves a sour taste in my mouth.

Also, the fact that it doesn't spend a lot of time at all disambiguating between stuff like machine learning, algorithmic feeds, LLMs, image labeling and other related systems, allows you to sneak rhetoric like “OoooOoooOooo AI capabilities are sCAaArY!” without even really talking about how, exactly.

It's kind of funny though that the bit where “Artificial Intelligence is gonna make new bio-weapons!!!!1” is placed, quite firmly in the “Evidence Gaps” (read: “Yeah, right.”) section, alongside with the whole “AI is gonna cause macroeconomic via increased productivity!!!!1 zomg!!!1”. I thought that was funny. Oh, huh. If AI capabilities are accelerating, I thought these topics would be a shoo-in. You'd have evidence of that, right? No? Hmm…

Infosys boss says vibe coding is no threat because there’s more to writing software than writing software

Translating from corporatese…

“Are you kidding? We're going to make so much money having to clean up and pick up the pieces from the idiots who let the hallucinating chat box do their jobs, we won't have enough people to take it all!!”

Health board apologizes for phishing staff with with bogus vacation day

What you're requesting is not something the security team can do.

Yes. Which is why the apology was, and should be, done by management. However:

Coffey added: "Cybersecurity education is important, but it needs to be done with judgment and respect. There are many ways to test phishing awareness without exploiting the very real stress, fatigue, and frustration healthcare workers are experiencing."

Johnson told reporters at a press conference that the test "missed a mark," and promised to investigate how it was allowed to be sent.

"What happened here, obviously, is that all the lenses that were required to review the scenario weren't placed on it," he said. "It's not reflective of how we value our employees."

That's corporatese for “we're planning to fuck the cybersecurity team over.”

Re: Being nasty to your own staff

Speaking as someone who works in security, I think it's likely the people who wrote that did not intend to be nasty and probably didn't even think that they were at the time. That doesn't speak well of them/us either, but it is different than the image you seem to have about their motivations or attitude.

I'm actually a lot more charitable towards the security tech folks here, because it's their job to find vulnerabilities, and they had found a doozy. One expects infosec folks to be adversarial. That's not the problem.

The problem is, having found that vulnerability doesn't mean that the proposed solutions that would be offered would be particularly useful:

1. Make the people who “failed” the test undergo more “training” — i.e., busywork. These people are overworked and nearing burnout. Making them stay back for a lesson that was basically “you should behave like you're not exhausted and harassed perpetually by insane workloads” is stupid, especially if it involves a 45-minute “training” session that the workers have to take in their own time on top of their insane workload.

2. Fire or subject them to disciplinary action. Congratulations, you've made it worse.

Again, the infosec people did nothing wrong here. They had found a social engineering loophole. Okay, it was one that anyone with eyes could see, but the solution for solving the problem was to make the people being exhausted and burned out not be burned out. And management doesn't want to do that. Anything but that. Won't anyone think of the rapacious neoliberals stripping public services to the bone stakeholders?

Like, two parties were getting fucked here — the workers, and the infosec professionals being thrown under the bus, when the blame should squarely be placed on management.

On the one hand, yeah… I can see how you'd think, as an IT security team responsible for ensuring that staff are resilient to phishing attacks, that this was the perfect vector to test user susceptibility to phishing.

On the other… well, the way to solve it is to ensure that the pressures that cause people to engage their System 1s and click that link is by improving the material conditions of the workers so that they aren't stressed and constantly worried about burnout that they are easy prey to these targets. Instead, the solution is… going to make things worse, because management are going to blame the workers for the problem, and either assign more “cybersecurity training” (i.e. more work) or disciplinary action (i.e. threaten to take their jobs away, thus increasing that material precarity).

Either way, you're still going to have that vulnerability, and may have made it worse. Good job, manglement!

Everyone hates frontier AI labs, says Palantir boss

"Condense" might be a better (if still inadequate) description.

Ted Chiang said it best, I guess.

with a touchingly human-like disregard for "correctness", "veracity", or anything like that.

Come now, give your fellow humans (except maybe that one guy) more credit. Humans know a thing to be untrue when they disregard “correctness” and “veracity”. People often intend to lie. LLMs do not. It's all values from a matrix transformation translated into whatever output value it's supposed to be. There is no intentionality, nor there is awareness, the way humans, or, frankly, any organism does it.

A system cannot disregard a thing if nothing about the thing influences how it responds.

Re: You Can Smell The Fear...

This presupposes that the “frontier AI companies”, who have essentially leveraged the same old trick with no further insight other than endless patch-jobs to prevent the same old vulnerabilities, while confusing the difference between “plausible” and “correct”.

Now, I'm not defending Palantir. The problem with Palantir is deeper and more fundamental, and maybe Karp is engaging in the old braggadocio and doing that good ol' smack-talking the competition because that's where he thinks the threat is, but like… he thinks his biggest threat is from the slop purveyors rather than, I don't know, a tribunal of crimes against humanity, and the existential question that every Panopticon owner has had to deal with since the beginning of time — is the problem the fact that you don't have enough force to suppress your serfs and punters, or the fact that your serfs and punters are perpetually at the edge of rebellion?

It's from the same gang of clowns who think that words are thoughts and text is cognition.

Translating it means, “we're going to chop up all your textual sources, and then feed all that paper-mâché slurry into something that can (mumble mumble mumble) and then make shareholders so much value. The most value. More value than you can shake a stick at! So much value you'd be sick of it!”

It assumes that all you need is “infrastructure” and that the basic idea of “artificial intelligence” is sound. never mind that it's never been a technically coherent term and an infinitely protean concept that can fit so many grifts business models…

lol, lmao.

maybe 'Clammy' Sam 'Stochastic Parrot' Altman and Dario 'Wario' Amadei can just say that the people who hate frontier models are doing it professionally, innit? or maybe the hate is, idk, enterprise-grade?

Why Amazon hates 'human-in-the-loop' AI governance

Everyone's focusing on error rates when the actual issue isn't that.

It's kind of telling that IBM brings up the warning on making sure that people are held accountable for management decisions, not computers, because that's what they said in 1979 as well.

And… yeah. That's the issue. LLMs are accountability dodges. The people forced to run LLMs are the moral crumple zones.

I find it notable that this quote kind of gives it away:

“If I sit down at my keyboard and I type a command that takes a service down, I caused an outage,” Brandwine explained. “If I run a script that takes a service down, it's still me that caused the outage. If my agent writes a script that they then run, and it causes an outage, that's still my responsibility.”

Those are the stakes. They're going to force you to accept those stakes, on the threat of unemployment. And then when something goes wrong (“Humans, like AI agents and systems, are non-deterministic.”), guess who gets the blame, and the sack?

Again, AI being used to squeeze workers. What a time to be alive.

It blocked us at 'hello!' Anthropic Fable 5 refusing innocuous prompts

Anthropic expects cyber defenders and critical infrastructure providers to use its Claude Mythos 5 model, which shares the underlying model of Fable 5 but without the same safeguards. Doing so, however, requires participating in the company's Project Glasswing program or the trusted access program that's being rolled out for select biology researchers.

And there it is. Model so big and scawy and the only way you can unleash it's full power is to join our totally super-s3kr1t club which has premium absolutely affordable pricing tiers.

Re: Yeah, but ...

we could use the data center ruins as a playground

Someone pointed out that data centers would make awesome laser tag facilities.

If your sex life is dead, you can blame Steve Jobs

Wait… what?

It’s pretty much undeniable that births fell faster in places with AT&T coverage

So that's a problem with AT&T coverage, not iPhones. Did they have data after the exclusivity deal that excluded AT&T coverage from the causal relationship?

Also, you know, correlation ≠ causation.