akersten

"How do I replace my social security card" with the guy metal detecting the beach, ostensibly searching for it, is pretty great too.

Xcancel before it got cancelled and I doubt anyone would have paid for it anyway

If my ridiculous $100+/m subscription to the single dumb pipe that has a government-endorsed monopoly in my area is anything to go by, they're extracting plenty of value from this customer at least.

It's useful because it let's me see the decisions the model will make before it wastes a ton of time implementing them. The model is smarter now but that doesn't solve for underspecification if it guesses my intent wrong

If your core service is getting more expensive to provide and competitors are busy eating your margins, why let someone else taste your secret sauce and only get paid for the tokens, when you can keep the good stuff (bio capability) for yourself, and net both the profit and the fame?

> When the unit doesn’t work correctly, the anger and RMA requests are directed back at the Raspberry Pi foundation.

They don't need the resellers help for that! I've already been burned twice thinking "I'll use a raspberry pi for this small project" and having it die within a month because reliably reading/writing to an SD card is hard or something.

Between the notorious unreliability, crazy price hikes, trademark or whatever disputes, and this anti consumer ewaste generating policy, I can't think of any reason I'd ever patronize them again.. my only regret is it took me two purchases to learn this lesson!

Their sample prompt takes me back a year or two:

> Find unused code in our web app. Exclude generated files and test fixtures, and check for indirect references before recommending a removal.

If I "/goal remove unused code" in Claude today, I would not even think to specify "check for indirect references" and "don't consider generated code dead code," those sorts of intuitions have been "built in" to the frontier models for a while now.

So I'm pretty confused by this, what does it do that I can't with an agent swarm?

Good point of precision, I read their site too quickly. I don't think my argument materially changes with that number instead however. There were over 40,000 students enrolled in my university alone. Generously assuming they only turn in one assignment per year, having 4 of them go through the "computer says you cheated, and as you can see, it's 99.9% accurate" gauntlet is not a price we should be willing to pay for... the marginal benefit of this tool over more classical ways to proctor and assess pupils

> I would have guessed that reliably identifying LLM generated text was not possible

It depends what you mean by "reliably." If you mean, "we should be comfortable relying on this kind of tool at scale to identify and punish students, professionals, and writers who may have used AI," absolutely not.

If you take "reliably" to mean "1 in 200 false positive rate" as they disclose on their front page, absolutely that is possible (they are doing it today!). If you think there are more than 200 assignments turned in over a given year at university, you probably do not consider a tool like this fit for purpose. It's an open question whether those procuring said tool are aware of this

Unfortunately their marketing is really insisting on the former, and trying to push it into the zeitgeist that detection of AI-generated or edited text is reliable-type-1 now and long-term. They fail to make it clear that this is merely a tool that strongly suggests text follows patterns known to us at the present time of known LLMs. However, that fingerprint will drift over time, as LLMs get better, human writing style evolves, and the line between human and "smart autocorrect" becomes even blurrier (does speech-to-text push the model into "AI assisted" mode, because it tidied up your punctuation, for example?)

"What color are your bits" is good reading today as it was 20 years ago: https://ansuz.sooke.bc.ca/entry/23

> maybe it’s better than nothing?

While that is not quite my bar of confidence when implementing wide-reaching technologies that have numerous unexplored knock-on effects, I guess the calculus must have been different on Infinite Loop recently.

The fundamental issue isn't technical. It's that people will see the "certified real" tag and just take the image for face value of whatever narrative someone wants to convey. They'll see the "Real Photo, Verified by Apple" and their brain will short circuit [0]

I don't think we should have this, for that reason alone (but many others too).

[0]: https://imgur.com/fVPkpuQ

Why are you subjecting yourself to that instead of downloading Firefox + uBlock origin?

See the headline of this article for Exhibit (I've actually run out of letters in the alphabet) why adblocking is a moral imperative.

We must ensure the gravy train keeps rolling until we IPO.

> Crack down on unauthorized distillation / prevent weight theft

Actually hilarious to put that in writing, given the genesis of this entire business model.

Eh, I hope Apple continues to provide this and it forces the needed discussion about how two-party consent requirements are nonsensical. Why should it be illegal for me to remember exactly a conversation that I participated in, instead of only being allowed to have a vague recollection?

Laws like this provide cover for abusers and deceivers, by preemptively spoiling objective evidence and making any accusations depend on hearsay instead.

> They were coordinating with OpenAI regarding a publishing timeline, but could not come to an agreement,

Skimming the PDFs it seems much more dramatic than that? It sounds like at least one of them is concerned OpenAI "solved" the problem by having their internal model use the chats of the independent researchers and want to claim the credit instead? I don't know. The tone is pretty accusational though:

> the one Levent and I had quietly chosen to attack. Almost nobody else I know of was working on it. It is not the direction one arrives at in a few days by giving a model the problem statement. When I heard “forced,” it was a bright red flag.

> I was shown a prompt and told the internal research model had simply been given the problem statement. Levent had been told by Sebastien “very little human input” had been used. This turned out not to be true. Over the course of the call, as members of their team sent Sebastien corrections and details over their internal chat, it emerged that an entire team had been working on the problem, that this was one of a number of things that was tried, that work had started on the unforced problem, that the team first set the model on easier problems, including Euler, that even the prompt that had been shown to me had been written by prompting Codex, and that an insane amount of compute had been used.

> I asked when the first prompt had been sent by them. This question was not answered directly by OpenAI for some time. Eventually it was agreed that it had been sent in the past few days, after information about our work had reached OpenAI.

> I asked whether the model had been trained on, or had access to, our sessions in Codex, into which we had been putting all our drafts for the whole of this project. I was told the model did not look up user data. I asked again, about training, and I did not get an answer. [0]

[0]: https://cims.nyu.edu/~tristanb/statement.pdf

There are only two types of scanned ID documents, those that are known to be compromised and those that are not

> So we take a web browser and trim it down to only ever show a single <textarea> element you say?

I was saying if the goal is "text editor," the web platform is the wrong foundation entirely

Ok, co.uk was perhaps a bad example, because it's owned by the same registry as the TLD, but perhaps there are other 2nd level TLDs where that is not the case. My point is both that it's hard to tell, and more broadly why would anyone want their domain to be tacked on to some 3rd level subscript anyway, when there's so many plain top level domains available. Surely most of us (present company excluded perhaps) do not feel so passionately about the reverence of `co.uk`

I don't have some nefarious desire to scare people away from the TLD of their choosing. Really I'm bringing it up to be like "why would you even, like, want some 3rd rate domain instead of getting a .com" so I don't think there's anything to correct

I know about the public suffix list - I was wondering about the wildcard specifically. In the very issue you linked to, as of 2025, it seems this was still unresolved...:

> We have no plans to modify the .name entries at this point in time. We are aware of the implications of adding a wildcard, therefore we won't.

It kind of seems like an insane TLD structure to begin with, right? I always thought .co.uk was bad (you're just pinning yourself to whoever owns the .co. part, but at least browsers have some suffix list where you can't, I don't know, hijack some login cookie for all of .co.).

Joe Smith and John Smith can independently register joe.smith.name and john.smith.name, do browsers have a wildcard suffix list for the 2nd level of `.name` specifically, or can Joe set a cookie on all of .smith.name?

What the author describes here is hard because it's "simple."

What's simple because it's "hard" is replacing parts 2 & 3 with a network appliance like TrueNAS running a zfs pool that syncs to backblaze every night. Yeah you have to learn a bit but it won't fall in weird ways like the hard drive part here will just fail to mount one night and not back things up for 3 months until you notice. My 2¢

Yes, who would have guessed that the <textarea> element, designed specifically for this use case and built into browsers for 3 decades, would be the most performant and behaviorally consistent way to implement editable text.

I'm kind of sad the author stopped shedding unneeded complexity there though... we're not really building a text editor yet, we're building a website with a fancy input field. If we want to build a proper text editor we must eschew the bloat that is the web browser too.

> Can someone explain what coordinated pacing is?

My charitable read is a legal cartel that allows the small club to switch to Marathon instead of Sprint mode, and drip feed us frontier models at inflated prices, while preventing open source and overseas labs from releasing models because they're unsafe (for the Blessèd Fews' profits).

My uncharitable read is somehow even less constructive..

> Also, a watermark doesn’t stop your tool from working for you. It just stops you from passing of its work as yours.

I think we fundamentally disagree on what "working for me" means, but I remain steadfast in saying we should not accept tools that have ulterior motives beyond producing the output desired of them by me, the user.

> Watermarking the outputs themselves is very different and much more effective compared to how tools like Pangram work.

At the end of the day the only artifact is text that you can do statistics on. It's the same problem as today, with the probability shifted slightly more in one direction. This does not assuage my concerns at all.

> they are incredibly unlikely with SynthID

I kept my commentary focused on text watermarking specifically because I agree, a synth ID image watermark false positive is highly improbable. There's plenty of noise to robustly hide whatever you like in an image. Text is simply too capital I Information-sparse and fragile.

> good faith watermarking attempts bad.

I would sooner call it "ignorant faith" (if they don't know what they are emboldening) or worse "don't care" faith (there will be false positives and they accept this to further some illustrious and arbitrary goal of Text Purity). Whether that be to prevent model collapse or help you not waste time arguing with bots online, to me the principled stance of "tools work for the user" wins..

> generated text being watermarked is universally good.

If it worked perfectly, maybe you could make this argument in a vacuum.

It does not work perfectly. (It cannot. It is by definition a heuristic). That means there will be false positives. There is a chance those false positives ruin someone's career. See [0] for just how easy it is to push SotA "AI text detectors" in one direction or another.

Now, with watermarks, instead of everyone to some extent understanding that AI text detectors are wishy washy woo, they are now Anthropic certified to detect an official AI watermark.

With that kind of false confidence in hand, the people who trust the "computer says you plagiarized" machine are never going to believe you when you say "it can make mistakes," they're just going to fire you/take away your scholarship/cancel your grant/...

This is all beside the fact that we should demand our tools work for us and not for some shadowy master. "Universally good," absolutely not.

[0]: https://freddiedeboer.substack.com/p/i-wouldnt-say-pangram-i...

> Clearly we can't do nothing,

Why is that so clear? I can think of hundreds of real daily problems I'd rather my legislators be focused on than deciding how old my kids have to be before they're allowed on some website.

I would like affordable groceries and public transport, not whatever this is.

If they really want to "do something" then require an 18+ ID at a cell phone points of sale. That'll be something that won't affect me and actually will temper a ton of underage Internet use. Then they can study the actual effect of that and whether the harms were exaggerated after all.

And, reminder, so no one here loses the plot: this is a backdoor feel good measure lobbied for by big tech to push us into a world with mandatory device attestation so their ad impressions are worth more. They could care less about your kids.

No, sometimes doing something really is worse than doing nothing.

I wish the author got more into the weeds of anitcheat and how it evolved over time. Jagex was banning bots based on mouse movement heuristics in like 2005. 20 years later and there's some JS framework on the front page here heralding that same kind of analysis as some revolutionary new approach. I can only imagine what the modern anticheat engine entails, especially in the era of LLMs that can reliably emulate a human player. (Some of them can even write quite technical blog posts that almost escape detection ;)

The brazen crime of asking for a resource and receiving it.

You have the power, and should exercise it, to rate limit bad actors

> People interpret business as hostility because they cannot conceive of the idea that they themselves should pay for a service.

If that's what the business really believed, they'd put up a paywall instead of literally giving away their content. Some businesses do believe that. Most of them realize it's more profitable to sell data about you instead. So screw 'em and screw this victim-blaming "well we wouldn't need all this invasive tech if it weren't for those pesky adblockers, you'll need a TPM to browse the internet because you didn't play along with my bad business model" nonsense. The internet existed before big adtech, and it will exist after. Long live free computing.

> Like if you don't want to be kicked around, you need to be the one kicking.

This is a punk stance but somehow the argument is "be punk and whip out the credit card?" My argument is "be punk and adblock."