post_below's recent activity
-
Comment on Why do we find Tildes (and Lobsters) more interesting than the alternatives? in ~tech
-
Comment on Digital Detox (formerly Screenless Day) discussion: October 2026 in ~life
post_below LinkI've never done a digital detox on purpose, but I often go backpacking or camping in places with no signal. Maybe the most interesting thing, for me, about going days without internet access...What "digital detox" steps have you tried in the past?
I've never done a digital detox on purpose, but I often go backpacking or camping in places with no signal.
Maybe the most interesting thing, for me, about going days without internet access (weeks if you add it up over a typical year) is that it's not a big deal. I don't think it's ever felt weird or uncomfortable.
Maybe that means I'm not psychologically dependent on being network connected. Or maybe it's that the profound connectedness of the natural world is a far better kind of connection.
Loosely related: If I'm talking to a human, I put my phone down and leave it there. It may not be a "detox" but I think it's valuable to intentionally not use devices in some amount of everyday situations.
-
Comment on Anthropic's IPO prospectus shows sweeping AI vision, surging costs in ~finance
post_below Link ParentNot sure what you mean, Amazon's first profitable year was 2003, Amazon S3 didn't launch until 2006. They'd be less of a behemoth without it, but far from fucked. Almost half of their profit came...Not sure what you mean, Amazon's first profitable year was 2003, Amazon S3 didn't launch until 2006.
They'd be less of a behemoth without it, but far from fucked. Almost half of their profit came from ecommerce related activity in 2025.
-
Comment on Visitor activity tracking for B2B in ~tech
post_below LinkIt sounds like all you want to do is see what people are doing on the site yes? In that case privacy concerns aren't as relevant as you might think. Standard request logging, which is already...It sounds like all you want to do is see what people are doing on the site yes?
In that case privacy concerns aren't as relevant as you might think. Standard request logging, which is already happening at virtually every site on the internet, already covers the majority of what you need. Essentially you need an identifier (ip address for guests, session ID for logged in users), the URI requested, and a timestamp. After that it's pretty simple code to construct a particular visitors path through the site. If you want to know what they requested via a form (like a search box) then you maybe need POST request logging too, though only if the results landing page doesn't include the term in the URL.
The only other thing you might consider adding is a JS beacon of some sort so you can more accurately tell when a session or pageview ends. With JS you can of course track other things, like scrolling and mouse/pointer activity, but that's not likely to be useful in your case.
The privacy concerns come in not from collecting basic visitor data, that's table stakes, but from storing it in a non-anonymized way for a long period of time. Especially if it can be linked to additional PII. You fix that by not storing it for very long. That way it limits the impact of a breach, and your ability to share the data with others. If you really want to influence the privacy practices of the company, get it in writing in a published data retention policy.
Note that you can store aggregate stats with identifiers stripped for as long as you want, no privacy concerns there.
The other potential privacy issue is profile building for use in marketing. If you're collecting data about users and then uploading it to, say, Google Ads for targeting purposes... well it may be legal but it means you don't actually care about privacy, you're funneling data to the digital overlords. Even if it's pseudo-anonymous, they have enough additional data to link it to individuals if they want to.
It's actually pretty easy to make a practical case against that: The minute you share data with a 3rd party you move into a new category under laws like the GDPR. Your liability goes up and the number of rules you have to conform to (and build for) under a host of laws in various jurisdictions go up too. A public privacy policy that excludes sharing data with third parties sidesteps all of that. And you don't lose much as long as you're not relying on third parties for things like analytics. The truth is ad targeting based on uploaded user data isn't really that effective. They push it because of an insatiable hunger for data, not because it's better than context or behavioral targeting in most cases.
If time allows, it's simple enough to build that it's worth considering rolling your own. You'll learn a lot!
-
Comment on Someone ‘torturing’ LLMs in a robot prison has triggered the dumbest debate in AI yet in ~tech
post_below Link ParentJust when I was thinking the popularity of the vibes motivated "ok so we understand the technology and that tells us that it's not conscious or alive but we don't understand everything so we can't...Just when I was thinking the popularity of the vibes motivated "ok so we understand the technology and that tells us that it's not conscious or alive but we don't understand everything so we can't know anything so maybe it's actually conscious, prove me wrong. You see? Gotcha!" is waning, someone comes along and rekindles it with a silly clickbait experiment.
As you say, it's essentially statistics. Tokens are not expressing actual feelings or sensations.
-
Comment on Anthropic's IPO prospectus shows sweeping AI vision, surging costs in ~finance
post_below Link ParentAmazon didn't make a net profit until 6 years after a (very successful) IPO they went into with heavy yearly losses. Their pitch was that they were building infrastructure and it would pay off...Amazon didn't make a net profit until 6 years after a (very successful) IPO they went into with heavy yearly losses. Their pitch was that they were building infrastructure and it would pay off later. They turned out to be right.
I don't think that's comparable to AI labs, it's hard to imagine their profits ever justifying their valuations. Once a warehouse is built, you have the warehouse forever and eventually your distribution network is good enough and you can dramatically reduce spending. With model training you can't ever stop and, so far, each new generation costs more to train than the last. And if a plateau happens the big labs are screwed. Like, dead in the water while everyone else catches up, prices tank and inference becomes a commodity.
But I'm not sure the market will agree with all of that. At least not right now.
In any case, Amazon is just the most famous example of a market trend that's been going on for some time now. Companies losing money and still having hugely profitable IPOs is pretty normal.
A lot will depend on perception though, right now the capital class believes the potential of AI is largely untapped and nearly unlimited, that could change. Anthropic is smart to IPO asap while the sentiment is positive.
-
Comment on Do you think you're special? in ~talk
post_below LinkEventually I lived long enough to have better context for how big a number 8 billion is. Realizing you're not particlarly important in the scheme of things is freeing, I highly recommend it. But...Probably a common feeling that many eventually "grow out of". So I guess the follow-up question is if that happened to you, what made you change
Eventually I lived long enough to have better context for how big a number 8 billion is.
Realizing you're not particlarly important in the scheme of things is freeing, I highly recommend it. But that truth doesn't take away from the importance you have to the people who matter in your life. That's entirely real.
-
Comment on The doomsday cult inside OpenAI in ~tech
post_below Link ParentIf you expand it to wealth and power consolidation, and capital otherwise corrupting systems and contributing to the destruction of our ecosystems, I'm right there with you. If it's just about...If you expand it to wealth and power consolidation, and capital otherwise corrupting systems and contributing to the destruction of our ecosystems, I'm right there with you.
If it's just about AI... It would be wonderful if that was the biggest of our problems. Or even a big enough part that erasing it could change the trajectory.
-
Comment on The doomsday cult inside OpenAI in ~tech
post_below Link ParentOk then! That's not true. It began work after training was complete. It was fully cooked. I don't know if it was cheaper, but it was definitely faster. Mathematics is ever so very different from...Ok then!
The model that helped solve the millenium prize problem began work four days after beginning training
That's not true. It began work after training was complete. It was fully cooked.
Compared to the decades of human PhDs working on this, it seems both faster and cheaper.
I don't know if it was cheaper, but it was definitely faster. Mathematics is ever so very different from recursive self improvement though. You changed the framing and then replied to your own version.
This is a weaker form of recursive self improvement, but it is one.
If that's what we're calling recursive self improvement, I don't think it's all that scary. No argument that agents are involved in various parts of the process, model labs have been talking about that for some time.
For LLMs to get out of hand they'd need to run the pre and post training process autonomously and accelerate it dramatically. Currently velocity is constrained to a significant degree by compute, so the agents would need to solve that first. Even if LLMs that smart and reliable existed (they absolutely do not), they'd still need to be given the resources to do it. It's like not these companies aren't paying attention to what's using the compute they're spending historically unprecedented amounts of money on.
If the other sort of recursive self improvement were achieved (the sort where a human just pressed go and the rest ran itself and resulted in a better model, faster, than a human led process), they'd say something. They would have just won one of the most important races in the field. It would be invaluable pre-IPO hype potentially worth billions of dollars. And, if it actually worked, they'd sell access to the resulting model (which would be both the smartest and most compute efficient model on earth). Not selling it would be equivalent to setting the 100's of millions+ they spent on training on fire. The biggest problem they'd have, with the best model on earth, would be figuring out how to handle the demand.
So we won't need to speculate, we'll be able to read the headlines. If they were smart they'd also announce a pause on recursive training while they assessed the risks. Otherwise governments would force them to take a beat just to quell the worldwide panic. People are freaking out about the possibility with no evidence that it can even be done, imagine the response to the thing itself.
It sounds extremely shortsighted and reckless to say "it's just hype, I'll panic when 10% of the world population is infected". Even 1% there would be far too late.
In this comparison, what is the pathogen? You'd need something self replicating and with a troubling mortality rate. Something observable. Otherwise it seems to me that you're just picking something scary and arbitrarily comparing it to the topic to inject emotion.
It sounds like what you're describing is how you feel about AI, which is fine, but you're framing it like a conversation about what is. So concrete, in fact, that by having a different perspective I'm shortsighted and reckless. Which is quite an accusation to make against an internet stranger.
-
Comment on The doomsday cult inside OpenAI in ~tech
post_below Link ParentI agree with this, but I think the situation is nuanced. The public mouth pieces of the big AI companies (Altman, Amodei, various engineers and project leads) have an established history of...Yea, I find it rather off putting that there is large group of people who seem rather quick to conclude that the people closest to this technology don't know it the best or can't be trusted to give observations
I agree with this, but I think the situation is nuanced. The public mouth pieces of the big AI companies (Altman, Amodei, various engineers and project leads) have an established history of predictions that turned out to be wrong, exaggerations and general hype.
It makes sense to question what they say, even if they genuinely believe it, it might still be wrong.
Personally I have yet to see anything close to the capability to be truly, autonomously, dangerous in the frontier models. They can do dangerous things when directed or misused or deployed badly, but so can a lot of technology.
The current wave of fear seems to be based in the idea that, any time now, they will achieve recursive self improvement and then they will make themselves dangerous really quickly. Except there's no evidence of recursive self improvement either. I mean, put an agent in charge of its own memory system and watch how long it takes before it writes bad judgement to the record and then compounds it by building more bad judgement on top, and and so on. If you give it enough time, it will make itself worse, not better.
That's of course not how recursive self improvement would work, instead it would be agents fully in charge of the training pipeline, but the point is that their judgement isn't good enough. They would make bad decisions, which would compound over time. They'd also make good ones, and maybe end up improving themselves on average, but not faster or cheaper than a human led process.
The doomsday fear all rests on the idea that this will change, and maybe it will, but we don't appear to be close to that yet.
It gets a lot of engagement though.
-
Comment on Jakob Ingebrigtsen pipped to victory by Dawit Seare in the men's 5km in a thrilling finish at the 2026 World Road Running Championships in Copenhagen in ~sports
post_below LinkNever head the term pipped before, turns out it's because it's regional: It comes from "pipped on/at the post" an idiom from horse racing. Or alternatively: Kind of a weird way to describe a human...Never head the term pipped before, turns out it's because it's regional:
It appears the expression is mostly used by those who live in England, Ireland, and Australia as nearly all of the published instances were found in newspapers and books from England, Ireland, and Australia.
It comes from "pipped on/at the post" an idiom from horse racing. Or alternatively:
In the 19th century Britain, to pip someone meant you wounded or killed that person, usually with a gun.
Kind of a weird way to describe a human running race but there you go.
-
Comment on Australian athlete apologizes after winning race despite involuntary bowel movement in ~sports
post_below Link ParentOP was up front about not knowing the details, I read it like "help me understand this" which the thread seems to have accomplished. For me as well, when I saw the headline earlier today I assumed...OP was up front about not knowing the details, I read it like "help me understand this" which the thread seems to have accomplished.
For me as well, when I saw the headline earlier today I assumed it was about running and thought it was kinda weird since people shit themselves in running competitions all the time. I also didn't care enough to look into it, but I appreciate the extra context in this thread.
-
Comment on Billing for web hosting in ~comp
post_below Link ParentThere are a bunch of uptime monitors with free tiers. I like HetrixTools, their free tier includes multiple geographic locations.There are a bunch of uptime monitors with free tiers. I like HetrixTools, their free tier includes multiple geographic locations.
-
Comment on Billing for web hosting in ~comp
post_below (edited )Link ParentBest of luck with it, if you run into issues post again! Side note: It's easy to switch people's sites to somewhere else if you get tired of hosting their digital real estate. It's not a forever...Best of luck with it, if you run into issues post again!
Side note: It's easy to switch people's sites to somewhere else if you get tired of hosting their digital real estate. It's not a forever commitment.
-
Comment on Billing for web hosting in ~comp
post_below Link ParentHosting is not cheap because of huge cloud platforms, hosting is cheap because hosting is cheap. It's not the result of market manipulation, the resource needs of the majority of sites are...Of course huge vertically integrated cloud platforms can be much cheaper and afford it to lure in people for free.
Hosting is not cheap because of huge cloud platforms, hosting is cheap because hosting is cheap. It's not the result of market manipulation, the resource needs of the majority of sites are genuinely tiny.
This is an industry decades into maturity, you don't need to be Amazon scale to offer cheap hosting. There are so many medium sized NOCs that have nothing to do with the big datacenter companies who will give you shared hosting for a couple dollars a month and a solid VPS for not much more. With far better uptime than you'll get from someone who's just getting started.
Meanwhile at small scale there simply can't be enough profit to justify engineer hourly rates. Server techs make much less than engineers hourly and a couple of experienced people can keep whole racks of servers running smoothly. Hosting makes profits via scale, each customer pays a few dollars a month and maintenance time is cheap when spread across all of those customers. Trying to force a few customers to pay for your server maintenance hours is unfair to both them and you.
-
Comment on Billing for web hosting in ~comp
post_below LinkReplying with this in mind: The web is full of places where you can host small (even non-static) sites for free or for a monthly fee so small it's akin to free. Because of that, your proposed...Replying with this in mind:
I want to work toward establishing good business practices, and that means charging people fairly - i.e. never simply because I can
The web is full of places where you can host small (even non-static) sites for free or for a monthly fee so small it's akin to free. Because of that, your proposed $10/month isn't even close to competitive. And a $50 setup fee! What is that paying for? Any competent web host has a provisioning script of some kind. It should take 30 seconds to setup an account, email, DNS, certificate, web space and permissions and maybe another 30 seconds for file uploads. If it takes longer than that, which is entirely understandable being new to hosting, that's not something the client should be paying for. Their monthly fee should cover everything, over time the host comes out ahead because most months a given hosting customer requires no time at all.
That wouldn't be true at first, because you'd be learning the intricacies of server management, but it's not the customers responsibility to pay for that. The comparison should be with the going rate for hosting, not your hourly rate for learning.
If I was working with someone and discovered they were paying $10/month for static site hosting, I would feel obligated to inform them that they're getting minorly ripped off. Minor because it's not a huge amount of money, but suspicious because it's many multiples of what they should be paying and therefore makes me question the motivations of the person charging them.
I'm not questioning your honesty, to be clear, just letting you know what I'd lean towards assuming if I had no context.
And I'm not saying don't host the site, it's useful to have full control over the server you're hosting client sites on, just don't try to make money from it. Charge something nominal or nothing at all. Your $10/month VPS could easily handle dozens of static sites with minimal ongoing hours invested on your part. Does it really make sense to charge each one of those $10/month when they could get equivalent hosting free, or for a dollar or two a month, somewhere else?
Web hosting is a mature, competitive and saturated market, banish all thought of it being a viable side business. The margins are too low. Instead, give your clients somewhere cheap and easy to put their sites and focus your energy on the other services you provide. If you're adding value on top of the hosting, charge for that separately and just throw in the hosting, amortized over a handful of clients it costs you basically nothing.
-
Comment on Anthropic's Dario Amodei calls for AI oversight, joined by Sam Altman and Elon Musk in ~tech
post_below Link ParentI think the coverage of distillation creates the wrong impression. To whatever degree it's happening, it's not as much of a cheat code as coverage (and internet commenters parroting other internet...I think the coverage of distillation creates the wrong impression.
To whatever degree it's happening, it's not as much of a cheat code as coverage (and internet commenters parroting other internet commenters) make it seem.
If you have the full weights of a model, you can very effectively use it to help improve another model (especially a smaller model). At this point it's safe to say that all the big labs are using their flagships to help train their smaller models in one way or another. That's not the kind of distillation people talk about though. The chinese labs don't have access to that.
Without the weights it's a very different situation. You have to try to distill by running prompts through the closed model and using the output for training. For a while you could get reasoning (COT) tokens from the models but none of the labs output COT unsummarized anymore. They either don't output it externally at all or they run it through a weaker model for summarization, which kills its training value. After that change, there have been some strategies to trick models into outputting COT but those holes have reportedly been closed.
In that scenario, without COT, distillation is far less valuable. It's still potentially useful but it will never give you anything close to parity. And it's less useful now that there isn't low hanging fruit to try to distill like reasoning and tool use behavior. These days all the labs have those basics covered and can train them in house much faster and cheaper than they could with distillation. Given enough compute of course.
If chinese labs are distilling they're doing it to polish their models after they've already done most of the training legitimately. Without a quality model to train in the first place this kind of blind distillation would accomplish nothing.
Unless the chinese labs have come up with new ways to prompt hack COT out of the models at scale (and somehow OAI and Anthropic aren't catching it) the most they're getting out of distillation is a bump. They definitely aren't able to press a "distill Astra" button and end up with a model that's just as good in X amount of time. There's no realistic value of X that could get them there without the target lab catching them.
-
Comment on Where do we go from here? (Regarding AI) in ~tech
post_below Link ParentAnother option, depending on how they feel about digital technology, would be to encourage them to learn about these new tools. Young, un-biased, minds thrive at the frontier. Traditional software...My eldest is on his last year of the high school equivalent here. He has a penchant for progamming and he really enjoys it, but I think I will advice him to go into some other proffession and keep it as a hobby.
Another option, depending on how they feel about digital technology, would be to encourage them to learn about these new tools. Young, un-biased, minds thrive at the frontier. Traditional software engineering is changing drastically, but using computers to solve problems is going to be a growth industry for a long time to come.
-
Comment on Best options for domain registrar and static site hosting in 2026? in ~comp
post_below Link ParentWow I wouldn't have guessed tucows was still around, they're from the early internet. I see they still inexplicably charge twice as much as they need to for domains :) To be fair, it used to be...Wow I wouldn't have guessed tucows was still around, they're from the early internet. I see they still inexplicably charge twice as much as they need to for domains :) To be fair, it used to be more like 4x.
-
Comment on Best options for domain registrar and static site hosting in 2026? in ~comp
post_below Link Parent+1 for Porkbun, they sell at close to cost and have a great backend. Stay away from Namecheap and Godaddy, they're well into the enshittification cycle.+1 for Porkbun, they sell at close to cost and have a great backend. Stay away from Namecheap and Godaddy, they're well into the enshittification cycle.
Or more accurately the Fediverse is what people, left to their own devices, make it.
And that's the problem. In a vacuum, the majority of people will tend towards being reasonably rational and kind. But it only takes a handful of people, in a given space, to change the vibe. Once people start getting triggered, reasonable and kind are the first things to go. It's just biology.
In addition to that, even a fairly low percentage of negativity and triggers is enough to put off most of the first timers you actually want to keep. So it becomes self selecting, over time you end up with a higher and higher percentage of people who are just there for the drama. Trolls and troll feeders essentially.
So any space that doesn't actively work to discourage people's worst instincts eventually succumbs to them. Not because people generally suck, but because people have built in systems that make them suck temporarily when they're in triggering environments.
The fediverse has the reputation it does for a reason and I think anyone who's interested in supporting and improving it should be taking it seriously, because what you're seeing is a selective process pushing away a lot of the people you probably want to stick around.
Which is why self moderation and "free speech" ethos have never really worked on the internet.