Greg's recent activity
-
Comment on Amazon shuts down internal AI leaderboard after employees cheated in ~tech
-
Comment on What internet discussion sites remain? in ~tech
Greg Link ParentYup! The only ones that immediately sprang to my mind are level1techs and war thunder forums, both of which are attached to larger IP elsewhere, and honestly I didn’t mention the latter because I...Yup! The only ones that immediately sprang to my mind are level1techs and war thunder forums, both of which are attached to larger IP elsewhere, and honestly I didn’t mention the latter because I only really know it from the memes about people leaking classified info.
There are vestiges of the real old internet out there too, but most of them are tiny enough that I wouldn’t know them to name drop, so I went for the first easy example (although now that I think about it, eFestivals is a good non-IP example that’s still got a bit of scale, as long as you don’t count a real world shared interest as an IP. And FlyerTalk just popped into my head as I was about to hit submit, too! There are a few still going now I stop to really think…).
-
Comment on What internet discussion sites remain? in ~tech
Greg Link ParentYeah, like you say, the only additions I can really think of are those very topic specific communities that manage to keep their own forums (e.g. level1techs), but those only really fit the bill...Yeah, like you say, the only additions I can really think of are those very topic specific communities that manage to keep their own forums (e.g. level1techs), but those only really fit the bill if one exists for your special interest of choice and you’re so into it that you’ll enjoy actually hanging out there rather than just dropping in once every six months to ask a question.
Although I do still love the feeling of running across an ancient but still active phpBB forum for something arcane I’m trying to learn from scratch, because that’s when you know you’re about to get an answer from the literal world expert in whatever topic…
-
Comment on The US Supreme Court's Republican appointees end civil rights redistricting protections in ~society
Greg Link ParentGenuine question, because I can’t remember the actual mechanics of this one: did they actually block the Democrats from acting, or did the dems just capitulate to threats?Genuine question, because I can’t remember the actual mechanics of this one: did they actually block the Democrats from acting, or did the dems just capitulate to threats?
-
Comment on It's not just X. It's Y. in ~humanities
Greg Link ParentI think I agree with the broad thrust of what you’re saying, but I also think there’s maybe a bit of an overestimate there on how much “LLM writing” is identifiable as a broad category. In my...I think I agree with the broad thrust of what you’re saying, but I also think there’s maybe a bit of an overestimate there on how much “LLM writing” is identifiable as a broad category. In my experience “ChatGPT writing” or “Claude writing” or whatever else are identifiable as examples of LLM writing, but the question I’d ask is how much that’s a technical consequence of current LLM capabilities and how much it’s a result of millions of people using the same baseline tool with subtle patterns that’ll show through as a result?
Most of the LLM output we’re seeing is the equivalent of McDonald’s food - cheap, mass produced, relatively low effort, identifiable not because it’s noticeably terrible in and of itself but because it’s the exact same not-great-but-fine product we’ve experienced many times before.
I think it’s maybe a bit of a trap to think we can identify LLM writing in general (and in turn perhaps let our guard down a little when we assume something is human written) just because we can identify writing from the most prevalent publicly available LLMs.
-
Comment on Why airlines are always going bankrupt in ~transport
Greg Link ParentI read @agentsquirrel’s points about pride in one’s work, employee happiness, and agents having, well, agency and ability and desire to achieve an outcome for the customer as the main thrust of...I read @agentsquirrel’s points about pride in one’s work, employee happiness, and agents having, well, agency and ability and desire to achieve an outcome for the customer as the main thrust of the conversation. That’s where I was coming from, but I can see that with a focus on different parts of that first reply, it reads differently.
-
Comment on Why airlines are always going bankrupt in ~transport
Greg Link ParentI deliberately didn’t address the “flight attendants should smile at me more” point in your first reply because I thought it was unnecessarily snarky, but since you’re doubling down on that side...I deliberately didn’t address the “flight attendants should smile at me more” point in your first reply because I thought it was unnecessarily snarky, but since you’re doubling down on that side of the question: I care about them being treated with respect as employees, as a valuable and integral part of a company rather than a costly resource to be bled dry. I care about customers being treated as people who are purchasing something in good faith and morally deserve to be provided a service in the spirit of that transaction, rather than as adversaries to be misled on pricing up to and slightly beyond the limits of the law, tied in paperwork to strip what legal rights do exist, and thrown into a kafkaesque nightmare of inaccessible and conflicting “support” when anything goes wrong. And I care about this as a pattern I see in effectively every industry, not remotely just airlines.
What is the social good in this case?
Maybe I wasn’t as clear as I tried to be, but I was taking this as a jumping off point for my thoughts on the erosion of the customer/service provider social contract in general, through the lens of airlines because it was the jumping off point that brought it to mind and I do think it makes for a good and relatable microcosm of the broader problem.
It’s why I was talking about craftsmanship, about ISPs and grocery stores and baseline economic competition in general, about holistic economic policy and its overlap with culture.
The social good is the rebalancing of the things we prioritise, to acknowledge that private profit is not the optimal goal, and should be treated as an imperfect proxy for actual contribution to the world.
In short: it was a train of thought, one that meandered beyond the relatively small station it began at. I think it’s fair to say there’s quite a bit more nuance in my replies than “plane tickets should be more expensive”, even if you do end up disagreeing with my broader underlying point.
-
Comment on Why airlines are always going bankrupt in ~transport
Greg Link ParentI’m really not talking about anything crazy like that - I might be looking towards the more pessimistic end there (not intentionally, but I’ll accept I might have rounded it that way), but it’s...I’m really not talking about anything crazy like that - I might be looking towards the more pessimistic end there (not intentionally, but I’ll accept I might have rounded it that way), but it’s not at all unusual to see £5k - £7k return tickets in business for transatlantic flights (the routes I’m by far the most familiar with, so again acknowledging my own bias there) where economy tickets on the same plane are going to be somewhere in the £450-800 range.
I don’t think I’ve ever actually bought a business class ticket - only ever miles upgrades or work expenses - and if I saw one as low as 3x the economy price on a flight I needed to take I’d really seriously consider it.
-
Comment on Why airlines are always going bankrupt in ~transport
Greg Link ParentI wouldn’t be averse to the government subsidies idea as a test case, at least in a world with better alignment of government to social good. Try it for everything from airlines to ISPs to grocery...I wouldn’t be averse to the government subsidies idea as a test case, at least in a world with better alignment of government to social good. Try it for everything from airlines to ISPs to grocery stores - but don’t give them any regulatory special treatment, and don’t subsidise existing profit making entities, just do the groundwork for organisations to exist that prioritise quality and service over profit. I’d be fascinated to see what would happen if every major industry had a cost neutral non profit option run in the public interest (it’d be counted as a subsidy because presumably the government would need to provide or underwrite capital investment, and accept the opportunity cost of not maximising return on that). Have the government fund a baseline array of services that compete with everyone on everything - force private companies to survive only if they can actually beat the nonprofit version in terms of what they offer.
That said, I’m already in kind of utopian territory compared to the current political reality. It’d probably need to be a cultural shift, and happen at a level far broader than airlines alone. As you very rightly say, the market isn’t some kind of divine force - but the strongly reinforced political and cultural zeitgeist has become to treat it like one. This succeeded in the market? It’s better. That failed in the market? It’s worse.
Success in the market solely means “convinced more people to pay for your product”. Maybe that’s because it’s a better product, or a better value despite being a worse product. Or maybe it’s advertising. Maybe it’s chemical addiction. Maybe it’s regulatory capture. Maybe it’s monopolistic behaviour, or a cartel. Maybe it’s externalising the costs and making them society’s problem.
A market is the study of human behaviour, and like any study it sets parameters on that behaviour, sets targets, shapes the very thing it’s studying. If you show that profit is the only criterion that matters, you get paperclip maximisers that increase profit incredibly well without regard to any other outcomes or externalities.
As a rough, very rough, back of the envelope idea, I’d suggest looking at an economic policy that still makes heavy use of market economics - markets are tools that do provably work for efficiently finding successful equilibria, after all - but couples it with a clear and intentional use of targeted interventions (regulations, taxation, direct state run competition) for the overall maximisation of net population happiness, comfort, and wellbeing. “This will hurt company profits” shouldn’t be an argument against a sensible intervention, it should be an acknowledgment that the companies involved were briefly allowed to gain from harmful behaviour but the loophole is now justifiably being closed.
And finally, to be clear: I know it’s not that simple. Government run entities are a conflict of interest when it comes to legislative special treatment. Regulation is subject to regulatory capture. Powerful and pervasive government intervention is susceptible to incompetent or malicious government. I’m not saying any of this is easy, maybe not even plausible. I’m certainly not saying one guy musing on the internet has the perfect answer. But I am saying we should be thinking and talking along these lines to see if there are ways to refine the practical details, rather than accepting the premise that market success is a sufficiently good proxy for what we actually want to encourage in the world.
-
Comment on Why airlines are always going bankrupt in ~transport
Greg Link ParentI know the market has spoken*, I just wish the old pricing and service were still an option. As with many, many industries, it’s been remodelled to support an incredibly price conscious consumer...I know the market has spoken*, I just wish the old pricing and service were still an option. As with many, many industries, it’s been remodelled to support an incredibly price conscious consumer at one end and and an extremely wealthy one at the other, rather than having a “decent price for decent service” option.
From the studies I can see, ticket prices in real terms are around half what they were per mile in the late 80s. I’m fortunate enough that I could and probably would pay double in a lot of scenarios for an all round better experience - not just a comfier chair or a better meal (although I fully would be paying for those too), but an experience that treats the customer and the employee both as people to be respected rather than as inconveniently ambulatory cargo and costs to be eliminated.
But I can’t pay double for that. I can pay 10x as much for business class, or I can pay a wildly varying amount for wildly varying quality in premium economy, but ultimately the standard path sucks and the good path is wildly expensive.
Same goes for a ton of industries: the default option is cheap and horrible for customers and employees alike. If you hint in any way that you’re driven by something beyond just minimising price, that’s a signal that you might be in the “wealthy” bucket and could be convinced to pay for a vastly more expensive boutique service. Decent quality for a fair increase in price simply isn’t an option in a lot of fields, where it used to be the case that it was the only option (which, yes, had its own issues of price floor).
And again, I know, the market has spoken. I still think it sucks, and I still think society as a whole loses out in far greater ways by deferring to the market alone. Maybe the value of good craftsmanship, or stable jobs, or even simply letting people run at less than burnout pace has been lost at a broad societal level in a prisoner’s dilemma of companies undercutting and consumers exercising rational self interest? Maybe we should be looking at ways to push the market into selecting for a more pleasant, more comfortable world for the people who live in it?
*caveats apply about how much markets can truly capture consumer preference when the options are limited to what providers offer and the barriers to compete are incredibly high. But yeah, I’ll accept that there is a genuine trend to cheaper-but-worse winning out.
-
Comment on Utah's shrinking lake: a scientific asset and a crisis in ~enviro
Greg Link ParentThe reasoning that lets environmentally damaging policies pass is pretty important to environmental topics more broadly, no? I’d expect it to come up whether that’s political, religious, or (in...The reasoning that lets environmentally damaging policies pass is pretty important to environmental topics more broadly, no? I’d expect it to come up whether that’s political, religious, or (in the case of the US in general and Utah in particular) a blend of both.
I’ll accept there’s a bit of snark in some replies, and I’m also biased as all hell on the topic, but I can’t really find it in myself to be that bothered by snark in the face of blinding hypocrisy. I think the broad point around “can we trust the environmental stewardship of people on record saying that God will fix it” is a crucially important one in a way you wouldn’t be hearing if it were, say, a Jesuit leader on record saying their faith-based position is that we should steward the planet to the best of our ability using the tools and knowledge at our disposal.
-
Comment on Utah's shrinking lake: a scientific asset and a crisis in ~enviro
Greg Link ParentNot OP, but it's not peculiar at all to me - I'd say even without there being explicit messaging it seems like a fairly understandable conclusion to draw by following the logic. I'm not saying...Not OP, but it's not peculiar at all to me - I'd say even without there being explicit messaging it seems like a fairly understandable conclusion to draw by following the logic. I'm not saying it's the only conclusion that makes sense, but it's certainly one that's come to my mind when I hear phrases like that.
For a related, and depressingly influential take that is explicitly stated in some fairly powerful denominations of American Christianity, just look at prosperity gospel. Some stop short and leave the implication of that one hanging, others are explicit in saying that those who aren't being materially rewarded are sinners or aren't strong enough in their faith.
I've heard from Indian friends that concepts like karma and reincarnation have also been (mis)used to justify hierarchy and discrimination on the basis that those at the bottom of the heap are there rightfully, thanks to transgressions either in this life or before.
Any just world hypothesis - and I'd say "God will never give you anything you can't handle" fits as a variant of that - will ultimately carry an implication that seemingly unjust outcomes are either the result of personal fault, or are opportunities in disguise, otherwise they'd necessarily be truly unjust and that doesn't fit the worldview. Which means if you put the worldview first, you end up in some very uncomfortable logical pits like the one @Omnicrola pointed out.
-
Comment on Utah's shrinking lake: a scientific asset and a crisis in ~enviro
Greg Link ParentIt's literal survivorship bias though. Most people make it through, but far too high a number don't - same as the "I smoked for 50 years / didn't vaccinate / didn't wear a seatbelt / etc. etc. and...It's literal survivorship bias though. Most people make it through, but far too high a number don't - same as the "I smoked for 50 years / didn't vaccinate / didn't wear a seatbelt / etc. etc. and I was fine" argument. You're only looking at the people who managed to find a way to keep going, and even then I'd say that more than a few end up breaking in a fundamental way rather than building as a person.
Through a religious lens, you can play the afterlife card to kinda sorta still make the thinking work. I don't like that as a fallback, but within the internal logic of religion it does hold together. Through a secular lens I don't personally see it.
-
Comment on What I learned building my first custom water loop in ~comp
Greg Link ParentIt's the difference between non-conductive (textbook) and non-conductive (real world). I'm going on pieced-together info plus what I half remember from university almost 20 years ago, but my...It's the difference between non-conductive (textbook) and non-conductive (real world). I'm going on pieced-together info plus what I half remember from university almost 20 years ago, but my understanding is that once you put a water based coolant into your loop it's going to start pulling ions from the metals it's in contact with and become at least a little bit conductive, regardless of how deionised it was to start with.
Account for the fact it's being physically agitated, held at slightly elevated temperature, and is expected to continue in that environment for months or years at a time and you end up with small effects that would normally be ignored by the spherical chickens in a vacuum model eventually becoming significant. Might well even be that it's still insulating enough not to damage a 12V circuit in case of a brief spill, but still conducts the... microamps? nanoamps? needed for slow corrosion over time.
As with a lot of things it's also a question of trade offs - you can mitigate the issue in a few different ways:
- Vacuum deposit a coating on any exposed metals that's thick enough to prevent direct contact with the coolant, and is safe from eroding over time, but won't impact heat transfer too much
- Use $5,000 of fluorinert as the coolant rather than anything water based
- Mix a chemical soup with enough additives to build up inert films on any exposed metal and/or soak up any free ions, without neutralising the biocide that's also in there to prevent algae gunking everything up, while remaining safe enough to use in the home, and without eating through plastic/rubber/etc. that you'll also find in a loop
And honestly I'd bet that any one of those is being used at least somewhere in a lab/industrial setting where they've got incredibly specific requirements. As I understand it some AIO manufacturers actually did find it worthwhile to go with option three when sourcing the parts at scale. But for a one-off custom loop it's easier to just avoid the issue!
-
Comment on What I learned building my first custom water loop in ~comp
Greg Link ParentHonestly I had no idea either before I tried it - that's partly why I figured it was worth getting the info down somewhere searchable. You're welcome, random person desperate enough to go to the...Honestly I had no idea either before I tried it - that's partly why I figured it was worth getting the info down somewhere searchable. You're welcome, random person desperate enough to go to the seventh page of google in 2032 ;)
And yeah, like 80% of my parts are Alphacool too and I've never had a problem there, so with you on that decision!
-
Comment on What I learned building my first custom water loop in ~comp
Greg Link ParentNice! Quick disconnects are incredibly worthwhile IMO, being able to isolate a component or part of the loop in the same way you'd unplug a power connector makes a huge difference to how likely...Nice! Quick disconnects are incredibly worthwhile IMO, being able to isolate a component or part of the loop in the same way you'd unplug a power connector makes a huge difference to how likely you are (or at least I am) to actually bother fixing and maintaining things.
The example wasn't hyperbole BTW, I do literally have that many blocks & disconnects in a single loop (server build) that I was expecting to need to pair with a more powerful industrial pump or multiple D5s - tested it with a single D5 to get a baseline and it's been running fine for over a year at this point, those things have way more headroom than any of the conversation I've seen elsewhere would suggest.
Strong recommend on Koolance QD3s, too, which I imagine you'll have seen mentioned basically everywhere. They really are built like they're designed to be the last line of defence sealing out a submarine hatch or something. Admittedly they're priced to match, but against the cost of components you really don't want water leaking onto I figure it makes sense. And their customer service team literally got someone at the factory in Korea to go out and Fedex an order to me directly when they were having freight issues, which got them a place on my vanishingly short "will shill for free on unrelated internet forums" list!
-
Comment on What I learned building my first custom water loop in ~comp
Greg LinkIf I could exemplary a post, I would! This is a great overview. It’s surprisingly tricky to get consistent information on custom loops if you haven’t built one before because the wider internet...If I could exemplary a post, I would! This is a great overview.
It’s surprisingly tricky to get consistent information on custom loops if you haven’t built one before because the wider internet has a lot of info that’s being parroted from the much jankier setups of 10-20 years ago and either isn’t relevant any more or we’ve just collectively learned more and realised it wasn’t the best advice even then.
I’ll throw in a couple that I’ve learned along the way:
- If you’re using a D5 pump, you probably don’t need to worry about flow rate. They can push through five blocks and 20 quick disconnects at 70% power no problem, but I see a fair number of people worrying that even a single minor restriction in a normal one or two block loop might cause problems.
- As long as you don’t mix materials with vastly different galvanic potential (i.e. don’t put aluminium in your loop), corrosion inhibitors are mostly for aesthetics. Most loops will have at least a couple of different metals (copper, nickel, brass, maybe stainless steel), so there will be some galvanic reaction - it almost definitely won’t burn through any functional parts, even over years, but it will leave a visible patina on shiny surfaces almost instantaneously. Inhibitors keep everything looking fresh and new (which is a totally legitimate reason to use them in the loop you’ve just spent hundreds on!), but they won’t stop functional corrosion with aluminium unless you use a terrifying concentration (which I’ve heard some AIOs actually do), and they’re almost never a strict necessity if you’ve got a loop with the other standard metals.
- Some premix fluids are nasty. I was getting blinding headaches after filling a loop with Aquatuning’s own brand stuff, turns out that’s a known possible side effect on the MSDS for one of the ingredients.
- If you’re going DIY fluid, distilled water with a tiny concentration (as in, parts per million) of cheap benzalkonium chloride is plenty to keep things clean and flowing. You can use that as-is, or add inhibitors or other additives from there if you like, and it lets you keep a closer eye on what you’ll be working with if that’s something that concerns you.
-
Comment on ‘It’s shameful’: New York’s elite lash out at Zohran Mamdani’s second-home tax in ~finance
Greg Link ParentProbably the most significant common factor I’ve noticed in the very wealthy people I’ve interacted with is an absolutely unshakable sense of entitlement. It’s interesting, in some ways I even...- Exemplary
Probably the most significant common factor I’ve noticed in the very wealthy people I’ve interacted with is an absolutely unshakable sense of entitlement. It’s interesting, in some ways I even kinda see positives in it as an approach - it’s helped me unlearn a lot of mostly implicit things I was taught growing up about being deferential, not taking up space, not making demands, generally putting oneself second to people who have no reason to be allowed to put themselves first.
The frustrating part is that only about a third of them seem to be… honest about that approach, I guess is the best way to put it? I can have a principled disagreement with someone who’s just openly trying to maximise results for themselves and expecting everyone else to do the same in a kinda sportsmanlike way. “You’d do the same in my position” - well no, actually, I try to be better than that, but I believe you’d expect me to and I believe you’d take it as a fair loss if I did.
What I can’t stand is the people who’ve somehow managed to convince themselves that they actually and uniquely deserve special treatment to an extent that they deny it’s special treatment at all. Those are the fuckers who are genuinely pissed off about this but are still willing to axe 1,000 jobs without feeling bad about it, and those are the ones whose mindset I’ve never been able to really get.
-
Comment on ‘It’s shameful’: New York’s elite lash out at Zohran Mamdani’s second-home tax in ~finance
Greg Link ParentThe real people line was the one that really got me, like what’s even their intended implication there? We’re real people, who are suffering the ultimate inhumanity of… paying slightly more for an...The real people line was the one that really got me, like what’s even their intended implication there? We’re real people, who are suffering the ultimate inhumanity of… paying slightly more for an enormous luxury asset that’s still going to appreciate in value? Do they genuinely believe that no other person has ever had to pay a bit more for something than they’d want to before?
-
Comment on How democratic governments came to view VPNs as circumvention software that must be restricted in ~tech
Greg Link ParentYou’ve got half the respondents giving an answer even on the porn sites question, and a vast majority of that half saying they wouldn’t provide verification themselves. Unlikely to verify beats...You’ve got half the respondents giving an answer even on the porn sites question, and a vast majority of that half saying they wouldn’t provide verification themselves. Unlikely to verify beats likely on every single category, from messages apps to pornography.
I think it’s a stretch to say that people actually support it in a meaningful way if they aren’t willing to obey the basic rules themselves. At best, they quite like the sound of the idea, or at least think they’re supposed to quite like the sound of the idea.
I also genuinely haven’t seen anything linking the OSA to weakening US tech. That might be my bubble, and I think I remember you mentioning that angle in another thread too, but for what it’s worth I’ve only ever seen that used around GDPR and similar privacy focused legislation.
I think that’s debating around the edges of the real issue, though. Like I said, the big question in my mind isn’t even “can you get a decent majority of people to think it’s broadly a good idea?”, it’s “why this, in particular, with such political forces and swiftness, when so many genuinely good ideas that are genuinely popular get nothing close?”
Funny how maximising to a stated goal without regard for intent or outcome is “cheating” when the employees do it but just good business sense when the companies do it.