The first time I saw a computer, I saw a machine for making things. I once read a quote from actor Noel Coward who said that television was "for appearing on, not watching", and I immediately connected it to my own relationship with computers.
I don't want an LLM to write software or blog posts for me, for the same reason I don't want to hire an intern to do that for me: I enjoy the process.
Everything else, I'm in agreement on. Writing software for yourself - and only for yourself - is a wonderful superpower. You can define the ergonomics for yourself. There's lots of things that make writing software a little painful when you're the only customer: UX learning curves flatten, security concerns diminish a little, subscription costs evaporate...
I actually consider the ability to write software for yourself a more profound and important right over anything the open source movement offers. Of course, I want an environment which makes that easier, so it's this that makes me more concerned about closed ecosystems.
LLM capability improvement is hitting a plateau with recent advancements mostly relying on accessing context locally (RAG), or remotely (MCP), with a lot of extra tokens (read: drinking water and energy), being spent prompting models for "reasoning". Foundation-wise, observed improvements are incremental, not exponential.
> able to do most programming tasks, with the limitation being our ability to specify the problem sufficiently well
We've spent 80 years trying to figure that out. I'm not sure why anyone would think we're going to crack this one anytime in the next few years.
> Foundation-wise, observed improvements are incremental, not exponential.
Incremental gains are fine. I suspect capability of models scales roughly as the logarithm of their training effort.
> (read: drinking water and energy)
Water is not much of a concern in most of the world. And you can cool without using water, if you need to. (And it doesn't have to be drinking water anyway.)
Yes, energy is a limiting factor. But the big sink is in training. And we are still getting more energy efficient. At least to reach any given capability level; of course in total we will be spending more and more energy to reach ever higher levels.
Books enter the public domain. Project Gutenberg and others produce DRM-free versions. Many academics and people who wish to share their knowledge also publish works DRM-free, sometimes under permissive (copyleft), licenses.
The fact you see DRM as the norm and non-DRM as “a unicorn” that “doesn’t exist”, is mildly sad. You should explore all of the above a lot more, and much more besides.
I assumed that that was clear from the context, but let me rephrase it then:
"being made available DRM-free on Amazon" (and I'd narrow that down to "primarily/only on Amazon")
Of course public domain books are DRM free but I'm getting those from Gutenberg, not Amazon. Likewise, the copyleft books I'll most likely download from their own homepages, not Amazon.
I'm aware that DRM free media exists, including for currently copyrighted content that Amazon distributes ;)
Mildly sad is also that you seem to fault GP for not “exploring” more, instead of the insane practice of DRMing everything in the first place.
I never have purchased DRM protected media and never will - I’d rather pirate everything digital and but physical hard copies.
I don’t actually think it’s their fault, and if they feel I’m faulting them, that wasn’t the intention.
I think it’s sad that what we thought everyone saw as a nonsense is now so normalised that alternatives are just disappearing from view. Everyone should be encouraged to explore.
Piracy is your preferred option, but when that became more mainstream we actually ended up creating the market for more DRM, in the form of iTunes, Spotify and others. I’m not sure I want the future of digital media to be entirely subscription-based like that.
What might be a better solution is showing that media creators can achieve more of their own objectives through releasing media without DRM. This only works if their objectives are not entirely around making money from media sales, and more aligned to influence, or audience building.
I’m actually surprised at this point that musicians - given they don’t make money from streaming services and see them as tools to build audiences for live tours where they really make their money - don’t just jump over already.
> Piracy is your preferred option, but when that became more mainstream we actually ended up creating the market for more DRM, in the form of iTunes, Spotify and others. I’m not sure I want the future of digital media to be entirely subscription-based like that.
Nah DRM and subscription models would have arrived even if there were no piracy. Subscription models because they guarantee income and most people use less than it's worth. And DRM because companies are paranoid.
In fact even now DRM serves no actual purpose other than harassing legitimate buyers. Pirates have no issue getting content and in the case of books they never will because copying can even be done by OCR.
I was just talking about books, but sure for music there are tons of alternative options as well.
I detest streaming platforms and it’s pretty easy to buy music directly from the creators in almost all cases - except maybe the top “superstars” but I would argue that they are probably doing fine anyway…
Also physical records still exist for music as well.
Lots of artists can do just fine with living from media sales.
Look I’m not saying “pirate everything and never pay the artists” - I’m saying “never pay the predatory tech companies that have inserted themselves between us and artists”
...and then they get re-packaged with DRM on Amazon's store, mostly because people uploading public domain books on Amazon have no idea what they're doing.
> Project Gutenberg and others produce DRM-free versions. Many academics and people who wish to share their knowledge also publish works DRM-free, sometimes under permissive (copyleft), licenses.
You can read DRM-free stuff on a Kindle already, so that's not particularly relevant here.
> The fact you see DRM as the norm and non-DRM as “a unicorn” that “doesn’t exist”, is mildly sad.
When every big publisher is doing it, it is the norm. That doesn't mean there doesn't exist any book publisher which doesn't do this, but the vast, vast majority of the books actually sold today contain DRM. We don't have to like that norm, but pretending it isn't one is just denying reality.
While lots public domain books are on Amazon's store, most of those books are not free, both in the sense of "free or charge" and "DRM free". A lot of literature classic are released by a major publishing house with foreword and annotations, which to be fair, are copyrighted works and provide value. And they cost a bit of money. The "real" public domain versions provide by Amazon are barebone. Those versions are often good enough for many people, but you don't need to get them from Amazon in the first place.
In other words, public domain or not does not have much to do with DRM-free or even Amazon.
Broadly true, but I think we can let the hardware-focused people in now and again, even if they do make an awful noise as they drag their knuckles along the floor :)
A great take, and thanks for your all hard work dang.
Yesterday the top comment on two stories I went to discuss had deep and meaningful content, before the last line which was a "and I talk about this stuff all the time on my newsletter [link]", and I was conflicted. Same poster each time.
The poster had done the HN thing: responded with thoughtful examination of TFA, unique and interesting insight, and I don't feel it was AI generated.
And then they marred it. They pushed something just slightly out of context. Not entirely, just a smidge.
I hope we can keep an eye on that sort of thing around here, it feels like it could slide into something...
The appropriate place to link to your website, newsletter, whatever is on your bio page (which people have to actively click into, specifically because they want to know more about you and potentially find such links).
I agree that linking to your own work in comments is generally bad form.
Yeah, this "btw I have a newsletter here" seems overly promotional. HN as a forum doesn't have support for "signatures", then it feels a bit off to end every post with something not really relevant.
The grey area is people constantly linking to their own blog, but the linked post is relevant (example [0]). Like, it's good when people post relevant links to diver deeper, but when it's constantly your own content, that irks me a bit.
simonw is a smart guy, definitely an HN darling for anything LLM related nowadays but at the same time he is constantly pushing for his personal brand, IMO. Maybe unconsciously and just because he is very prolific but still, I get that feeling.
He writes because he loves writing. People can tell, and thus like reading it. If that creates a "brand", there's not much he can do about it. He's an active, valuable, good-faith participant here, has a subject matter expertise in something HN obviously cares very deeply about, and isn't going anywhere.
I'm not denying that, in fact I think it's basically because he writes a lot and participates a lot. But yet, it can give the feeling that he is curating his own brand. But probably it's just an organic thing, not a fabricated one.
I don't think I have been rude at all, and also seniority does not give any special authority over opinions.
My point was that the "creating the personal brand" part - which can seem something intentional - is just a byproduct of posting, writing and contributing. And he contributes because he has knowledge, opinions and things to say.
It's like writing your own blog with good content and getting organic traffic from search engines vs writing SEO content to get traffic and get noted.
Maybe I initially expressed myself not the way I really wanted.
I'll be honest and blunt: I try to avoid his blog posts and comments as much as possible because I do find his contributions to be super spammy. It's about the frequency of his self-promotion (as a researcher in NLP it's tiresome to constantly see his self-promotion on nearly all posts on HN related to LLMs). Seems I'm not the only one.
If he posted on the ML subreddit while I was still a mod there (left after the API kerfuffle) I would have messaged him and asked him politely to tone it down.
I have no issue with simonw's work/comments; I've never gone near it because it's far afield from my non-tech world. He is, though, someone who breaks one HN guideline:
>Please don't use HN primarily for promotion. It's ok to post your own stuff part of the time, but the primary use of the site should be for curiosity.
His submissions here are usually to his own work. Admittedly, that HN guideline is like an obscure 19th-century law which is still on the books that few know of and is never enforced. Even so, he's clearly well-regarded based on the amount of conversation his blog posts elicit.
You're a prolific commenter here, so please don't play coy. He's constantly linking to his blog and llm tool.
Clearly you feel differently and since your MO seems to be that you always wants the last word, have at it. I try not to be terminally online, so I'm not willing to engage further.
His LLM tool rules. I cite it too. Used it a bunch earlier this year for the municipal data work I was donig. Why wouldn't he cite it? Avoiding these kinds of mean-spirited criticisms would require him to twist himself into a pretzel. I'd rather we just all agree not to say shit like this.
> Yesterday the top comment on two stories I went to discuss had deep and meaningful content, before the last line which was a "and I talk about this stuff all the time on my newsletter [link]", and I was conflicted. Same poster each time.
In my opinion, if they put thought into a deep and meaningful answer, then I think that's fine. If they say "Oh yeah I talked about this in my blog [link]" that's totally different.
It's really contextual so... maybe? A good comment that just happens to do it once, likely not. But OP saw it multiple times in a short time from the same user, so probably.
Constant self promotion is what is hurting so many other platforms, I'd hate to see this one take even a step in that direction.
If it's only a link to a newsletter, sure. And "signing" a comment breaks a norm here. But if it's otherwise a good comment, seems like a poor use of "flag"; one of the rare instances where maybe you want to leave a short, polite note ("don't sign off with links to your newsletter, we don't sign things here").
Flagging deprives everyone else of the comment, and is especially hostile if there's already a thread sprouting from that comment.
I have a different take. If someone uses effort to post something genuinely interesting just so they can advertise their blog, good. You don't have to click the link, and it's better than nothing (assuming other comments here aren't as interesting).
The fact that HN does not work in exactly the same way as the horrid platforms out there that automagically surface things that make neurone in your head scream "click that thing! click it! look! LOOOOOOK!", does not mean your brain doesn't treat it the same way.
This isn't unique to this technology: books with interesting titles and covers sell better than books with boring titles and covers, even if the latter has more interesting content.
I think what makes HN a little different is that a lot of places we might congregate online only exist for the flamebait, or are specifically built around engagement metrics to serve DAU/MAU numbers and advertising asks. You just happen to take a bit more chance on here than you would anywhere else, and that's absolutely 100% a good thing.
If there are controllable inputs that skew you to one side of the variance, I hope the moment someone discovers them, they are shut down - otherwise this place just becomes another hell-hole.
I have the impression that there's a dashboard that the admins use for "second chance" to allow them to schedule stories they think didn't do as well as expected, to give those stories a second (third?) bite at the cherry.
One sometimes sees suggestions from @dang to resubmit, and/or comments about giving stories another go around if they did worse than expected.
The only other manual intervention at that level I've heard of is to promote Ycombinator companies; but I'm not sure what complexion that takes.
There's possibly some manual actions around vouchs/flags, to revive comments that got flagged and have now been vouched for. Maybe it's all automated but I suspect there's judgement applied there; just based on how the site is run in general.
No particular insight, just been around here for ~15 years (eek!).
Couldn't agree with you more. The internet has, to borrow the word of the moment, become enshittified, and people think that's "normal".
People accept that platforms should be centralised, and that they should harvest your data in order to sell it to adtech companies who will then feed it to an industry that learns in real-time how to prey on your darkest fears to sell you things you don't need but might make you feel slightly less sad for a second. And people just accept it: that's normal these days.
They even call it doom-scrolling, and don't ask "wait, should I want to scroll through actual doom? Is the occasional video that makes me smile really worth it all?"
Perhaps it's my age, but I can't understand anybody who says their main form of media consumption is YouTube. How? How do you actually put up with that, knowing what is behind every mouse movement and click, and the knowledge that every single pixel in front of you is being tweaked by robotic neuroscientists squeezing every drop out of A/B tests to make you feel like utter crap? Like, seriously, WTAF?
HN is popular within its niche precisely because it isn't like that. It is not "a platform", in the modern and now normalised sense. It links out to other sites and asks people to come back together to discuss what they saw there. Old school. No ad tracking. No doom scrolling. Pick what you like. Click it, don't click it. Discuss it, don't discuss it. Nobody is tracking "engagement". There's some gamification, but does anyone _really_ care?
This type of interaction is entirely native to my generation and older (I just squeak into millennial, on the older side), but feels completely bonkers to people who think Facebook, Instagram and TikTok are what is normal and how the Internet works.
Some of know they're not normal. We know they're aberrations, ghouls that prey on unwitting masses.
YouTube is easily the best “social network” in terms of having high-quality content with minimal manipulation. And I definitely consider it my main form of media consumption.
You just have to curate your feed and add stuff to playlists, not watch whatever is on the logged-out default home page.
What? You have to do a lot of curation and fighting the service itself (hiding shorts) to get to a bearable level. And I did not spend hours clicking things I don't like. I mostly just watch movies/clips I want, but my recommendations are full of clickbait manipulative garbage. So yes, one can find jewels in it - but only if digging through the garbage first.
There's actually a bit of crossover. This paper is quite old, but I know that other teams and other surgeons have visited each other plenty since this was published, and there's this cross fertilisation of ideas.
F1 is "important" in the sense that it is competitive, and so teams want to iterate and improve constantly. I think the fastest pit stop in the 2025 season 1.91 seconds, in which: the car is jacked, four tyres are removed, four new tyres are placed and secured into place, the car is dropped, the lane is checked for traffic, and then the car can move. There are thousands of permutations of how to get this right and that fast. And accuracy is important: get it wrong there is a risk of injury at worst, or a fine for an unsafe release at best.
ICU is obviously important in a different way. You can't really "experiment". Iteration needs data. So you need to go out and learn what good looks like from different disciplines, and then carefully plan the changes you want to make and get buy-in. Get it wrong, and people die. Best case scenario you're struck off, worst case you're going to prison for murder.
In dev speak, F1 can afford to be agile, ICUs need to be waterfall.
But because F1 needs to be precise and they perceive the dangers of imprecision so acutely from a monetary perspective (where you finish in the Worldwide Constructors Championship directly affects the profitability and viability of the team), they want to borrow ideas too.
It sounds ridiculous that surgeons and F1 garages would have so much to talk about, but it turns out, they really do feed ideas off each other sometimes.
> ICU is obviously important in a different way. You can't really "experiment". Iteration needs data. So you need to go out and learn what good looks like from different disciplines, and then carefully plan the changes you want to make and get buy-in.
I would think test runs with simulated patients offer plenty of opportunity to experiment.
> Get it wrong, and people die. Best case scenario you're struck off, worst case you're going to prison for murder.
Get it right and people may still die. The whole reason for the improvement effort is that the current practice is excessively risky. No one is getting fired, nonetheless going to prison for trying a sensible improvement to reduce the odds of a child dying which they were approved to attempt.
The regulation is not some bureaucratic nonsense, it exists for a reason.
If its existence means the manufacturer has to do more work, and the cost might rise, that is considered worth it.
They can lobby to have the regulation changed, and regulations have changed in the past based on industry input, but saying "it's too hard" does not make the product suddenly legal.
The first time I saw a computer, I saw a machine for making things. I once read a quote from actor Noel Coward who said that television was "for appearing on, not watching", and I immediately connected it to my own relationship with computers.
I don't want an LLM to write software or blog posts for me, for the same reason I don't want to hire an intern to do that for me: I enjoy the process.
Everything else, I'm in agreement on. Writing software for yourself - and only for yourself - is a wonderful superpower. You can define the ergonomics for yourself. There's lots of things that make writing software a little painful when you're the only customer: UX learning curves flatten, security concerns diminish a little, subscription costs evaporate...
I actually consider the ability to write software for yourself a more profound and important right over anything the open source movement offers. Of course, I want an environment which makes that easier, so it's this that makes me more concerned about closed ecosystems.
reply