1st: the presumption that AI generated text is actually unsuccessful, rather than proliferating broadly unchallenged today
2nd: the disposition that negative attitudes towards AI text are unjustified discrirmination, rather than working as a strong latent predictor of low-effort content
3rd: the assumption that human writing is reliably doled benefits, rather than some poor proxy of it (winning the social contest for claiming authenticity)
Operationally, only a very small minority of humans actually successfully identify AI generated text at rates ≥ Pangram. People discriminate against the label of "AI", but mostly fail to vote accurately. It's not uncommon to see bots abusing this gap for their own success -- accusing humans, sympathizing with generated profiles... FUD environment where people routinely get away with dismissing true accusations.
For someone who is mediocre at detection, this would structurally feel like an unhinged, unjustified bias: look at all these good posts, these honest people, getting undermined by discrimination...
Interesting points! though I don't follow them all.
> 1st: the presumption that AI generated text is actually unsuccessful, rather than proliferating broadly unchallenged today
I can only talk about HN. If you think this is proliferating broadly on HN itself, I'd like to see such links. Assuming it's "broadly", they should be easy to find.
> 2nd: the disposition that negative attitudes towards AI text are unjustified discrirmination, rather than working as a strong latent predictor of low-effort content
I don't understand this bit.
> 3rd: the assumption that human writing is reliably doled benefits, rather than some poor proxy of it (winning the social contest for claiming authenticity)
Sorry, but I don't understand this either.
> only a very small minority of humans actually successfully identify AI generated text at rates ≥ Pangram.
That's not what we seem to be seeing. I do agree that there's a wide spectrum and a lot of wrong guesses.
> It's not uncommon to see bots abusing this gap for their own success
I'd like to see specific links of this on HN itself. Again, if it's not uncommon, those should be easy to find.
What about people who genuinely write their own articles but due to their time spent with AI, they are sounding like AI? I know people like this. They really are writing like that now. Shouldn't there be fairness to them?
I do hear this sometimes but it's not clear to me how prevalent this is. I'd need to see specific examples.
Edit: these are just gut-feeling numbers but we've had many dozens (perhaps hundreds) of email exchanges with users about this to date. Usually when we ask them if their post was genai, probably 90% say "oh yes, sorry, I didn't realize you had that rule". Of the other 10% or so, when we ask further, it usually turns out that they used some other tool (a grammar checker, a machine translation, some kind of toucher-upper, etc.) which left LLM imprints on their post. It's pretty rare for someone to insist that they wrote the text entirely by themselves with no LLM or similar tool. Of those, I'm sure some are false positives.
There have also been a few instances where I didn't fully believe them, mainly because they'd already posted other things that got classified as genai. Fortunately this is rare. The vast majority of users who we have this sort of conversation with are sincere, legit users who don't know much about HN. Our hope is therefore to convert them into informed community members!
Most parsimonious explanation IMV: site staff can't see most AI slop. Reasons unimportant, but moderation systems are guaranteed to break down when the moderators themselves have poor classification ability.
A simple beneficial step that would lead to modest improvements and little downside: partner with Pangram. Either adding it as an automated spam filter, or by simply attaching the detection % to all posts.
what's the novelty? mixing healthcare together with a spa is an idea older than Christ. USCT is decades old.
Their butterfly chips might be cool, but it's not like the article says anything about that. There's only one other comment in the whole thread that even mentions it.
I find using tens of thousands of ultrasonic chips, submerged underwater to provide you a radiation free full body scan, all while processing a petabyte of data per scan a pretty ambitious and cool project. I hope they make it work.
1 petabyte per 60 second scans implies a kind of comical data rate to storage, even at RAM speeds that’s implausible. Imagine we need to write these to hard drives, they happily sustain 150Mb/s on the high end, which would imply you’d need 115,000 hard drives to absorb that amount of writes. Even with top end NVMe drives you’d need a thousand of them writing simultaneously.
That's likely the datarate of the ADC chips. You would downsample them directly on the FPGA board and maybe perform an FFT or similar transform. 16 TB/s across a few dozen FPGA boards is nothing crazy. After some early stages in the signal processing you might transfer 1 or 2 TB/s over ethernet to the servers. Entirely feasible considering we have 800 gigabit/s ethernet.
You’re completely right, this is why currently ultrasound reconstruction happens on FPGAs. They would need a lot of them given the number of transducers.
https://pmc.ncbi.nlm.nih.gov/articles/PMC6057541/
Yeah, and the author is obviously lying. Ever since mid-2025 it has been a winning move for AI users to categorically deny all allegations & reap the labor rewards of generated content.
I read the suggestion as being that the world’s poor aren’t paying for AI tokens (or subscriptions) to generate software from prompts, and that sharing the _output_ is more beneficial to people. The reduction in energy waste is secondary, and benefits _everyone_.
Have you run a system in production? There are a multitude of reasons that a system can go down. There's no indication so far from Anthropic that this was merely compute limitations.
> There are a multitude of reasons that a system can go down.
Start doing post mortems then!
At the very least, them using any off the shelf service that's shitting the bed would inform others to stay away from it - like an IAM solution, or maybe a particular DB in a specific configuration backing whatever they've written, or a given architecture for a given scale.
Right now it's completely like a black box that sometimes goes down and we don't get much information about why it's so much less stable than other options (hey, if they just came out and said "We're growing 10x faster than we anticipated and system X, Y and Z are not architected for that." that'd also be useful signal).
Or, who knows, maybe it's just bad deploys - seems like it's back for me and claude.ai UI looks a bit different hmmm.
I have no inside knowledge of Anthropic. But having done a lot of postmortems in general, one of the key dynamics that routinely comes up is "we know we keep shipping breakages, and we know these new procedures would prevent many of them, but then we wouldn't be able to deliver new stuff so quickly". Given where Anthropic is at and what they believe about the future of software development, that's a tradeoff that they may very well be intentionally not making.
Yeah, this is not just inference. First thing for me was an MCP I use went down in Claude Code, models still worked. Now "API Error: 529 Authentication service is temporarily unavailable."
HN is for anything that gratifies intellectual curiosity: https://news.ycombinator.com/newsguidelines.html. Historical and/or unexpected materials are welcome here! Having them on the site is a long tradition. (As is the "why is this on HN" comment, of course.)
It sounds like you know your Orwell - want to share something about that?
Hacker News Guidelines
What to Submit
On-Topic: Anything that good hackers would find interesting. That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity.
Haven't read the book, but points two and three definitely struck some bells in the back clocktowers of my mind.
More generally, reading a bit of Orwell was inescapable in my schooling, but I sought out 1984 myself. I discovered I had kind of a thing for both utopias and dystopias.
And as I contemplate things I might write or compose, I do note that outrage towards this regime is very much in the mix of my motivations.
if it does so happen that the crash originates from a browser exploit, you should expect to be more at risk due to the absence of a crash on an older version, not less
1st: the presumption that AI generated text is actually unsuccessful, rather than proliferating broadly unchallenged today
2nd: the disposition that negative attitudes towards AI text are unjustified discrirmination, rather than working as a strong latent predictor of low-effort content
3rd: the assumption that human writing is reliably doled benefits, rather than some poor proxy of it (winning the social contest for claiming authenticity)
Operationally, only a very small minority of humans actually successfully identify AI generated text at rates ≥ Pangram. People discriminate against the label of "AI", but mostly fail to vote accurately. It's not uncommon to see bots abusing this gap for their own success -- accusing humans, sympathizing with generated profiles... FUD environment where people routinely get away with dismissing true accusations.
For someone who is mediocre at detection, this would structurally feel like an unhinged, unjustified bias: look at all these good posts, these honest people, getting undermined by discrimination...