HN user
Former historian, current software developer.
https://ellen.dev/
Email: hn at my domain
The person who runs archive.today decided to involve me, and every other visitor, in their dispute. They decided to use us to hurt someone else. That's a pretty big sin in my book.
To work, you need to provide a National Insurance number, which is unique and tied to certain state benefits like pension. The idea is you work, pay "national insurance" contributions and accrue "contributing years" to get a state pension later.
The wrinkle is that it doesn't seem to be tied well to identity. Someone working illegally can provide an NI number that's legit but not theirs. Their work accrues to someone else's NI record, but the person getting the extra years probably never notices and the person working under their NI number doesn't care because they aren't entitled to a state pension anyway, they just want to work now.
I've been thinking about whether "$some_country rape gangs" seems racist to me. I've come down on "yes".
The reason might seem odd. But it ocurred to me that if you want to use immigration to reduce crime, including rape, the obvious solution is to ban all male immigration.
That shocked me because it seems so wildly discriminatory. Yes, most violent crimes are committed by men. But very few men commit violent crimes. Banning male immigration would punish a large group for the appalling actions of a few. Making it about "$some_country's men" doesn't seem a whole lot better. It's still unjust to punish someone for someone else's crime.
If anyone is curious about the exercise, I recommend trying it. It was disconcerting to sit with the idea of banning male immigration, really seriously consider it and realise how viscerally shocked I was by the idea.
Edit: for context, in the UK right now, phrases like "rape gangs" are part of the debate/argument about immigration.
"Approximately 1 million a year net" is completely true. In 2023 the UK's net migration was 906,000. But for context I'll add that it was a historic high and in 2024 net migration fell to 431,000.
The Migration Observatory at Oxford publishes excellent summaries about migration trends, e.g. https://migrationobservatory.ox.ac.uk/resources/briefings/lo...
Based on "a third of customer bills go on interest payments", I'd guess the original comment was about Thames Water.
In 2023 their interest payments were 28% of revenue. They also made the news for dumping particularly large volumes of sewage into rivers.
https://www.theguardian.com/money/2023/dec/18/water-firms-us...
Love a bit of source analysis.
I'd widen the frame a bit. People scared of losing their jobs might underestimate the usefulness of AI. Makes sense to me, it's the comforting belief. Worth keeping in mind while reading articles sceptical of AI.
But there's another side to this conversation: the people whose writing is pro AI. What's motivating them? What's worth keeping in mind while reading that writing?
what you’re demanding is that Apple allow you to run your own software on their OS
Yes. I'm not the original commenter, but this is what I expect.
From my POV, the OS exists to virtualise the hardware it runs on. I don't want the OS manufacturer to decide if I'm allowed to have a web browser or play games.
Naive in hindsight, but until game consoles and smartphones came along, it didn't occur to me that an OS would forbid me from installing something.
Admittedly a minor point of interest, but the last famine in England happened in 1623 and was local to an area called Westmorland [0]. That was 150 years before the Industrial Revolution, so the 10% figure might not be very reliable.
Channels about autism also disproportionally cover people who are willing to talk about their autism. Recently I've been reading The Lost Girls of Autism. Something that stood out in the anonymous accounts is the fear of being "discovered" and the associated anxiety and depression. Since reading that I'm not super comfortable with the idea of incentivising high-functioning people to hide.
I don't know if voters and non-voters have the same political leanings. It isn't something I've ever looked into. My observation was merely that measures of statical confidence assume random samples. Extrapolating from a non-random sample can give odd results. But this isn't a research paper, so it doesn't much matter.
Under that rule of thumb, anything accused of being an MLM will seem like an MLM, even if the accusation is absurd.
"Uh, what?? Air isn't an MLM, it's just the thing we breath. We all need it to live."
"Huh, I guess air is an MLM after all. Wild."
The electorate self-selected into voters and non-voters, it wasn't a random sample. Some chose to go to the polls and some chose to stay at home. Voter preferences don't say a lot about the preferences of non-voters, who've already shown they choose differently.
I've been mostly happy with a Terrmaster DAS attached to a mini PC running UnRaid. The bays are hotswappable and overall it's been solid.
I say "mostly" happy because I almost returned it. The USB connection between mini PC and Terramaster would be fine for a few days and then during intense operations like parity checks would disconnect and look like parity error/disk failure, except the disks were fine. Eventually realised the DAS uses power from the USB port as well as the adapter plug and the mini PC wasn't supplying enough power. Since attaching a powered USB hub it's been perfect.
Explanation of symptoms and solution, in case anyone is considering one or has the same problem: https://forum.terra-master.com/en/viewtopic.php?t=5830
In the study I'm aware of, 18 patients became infected in hospital and there was a 28% death rate within 30 days [0]
[0] https://www.journalofinfection.com/article/S0163-4453(16)301...
From what I've seen and heard from others, budget shenanigans and office politics are both common reasons for hiring contractors/freelancers.
If it's September and your budget expires in April (new tax year here in UK), hire a few contractors to make sure you spend all that money. After all, you don't know what next year's budget will look like. And not using your budget might be taken as proof you don't need as much money next year. So hire contractors and do that project now, rather than waiting and taking the risk you can't do it all.
For even more budget shenanigans and org politics, imagine you're a product/project/whatever manager in charge of an exciting new project with a big budget. You do not control any of the dev teams. For some reason the dev leads are not excited by your new project and are reluctant to assign people to the project. They tell you it will be next year and you wonder if they mean never. Luckily, you have a big budget for the project! So you hire contractors and now your project has a dev team. The company doesn't have to keep on the temporary team and the project is delivered. Success!
This scenario kicks several massive cans of worms down the road. But for a while everyone's happy while progress is made on the project. Later the dev teams will be annoyed when they have to maintain code they didn't want in the first place. And whatever organisational problem meant the dev teams didn't want to get involved will still exist. But, by hiring contractors, everyone gets to ignore those problems for a while longer.
It seems more like a contract dispute to me. My first thought is that a commercial court could decide who keeps the money.
(Clearly in the real world the American authorities have decided it's a criminal matter. I just find it odd that a "smart contract" dispute falls under criminal rather than commercial law.)
I don't know where I stand on this, but not everyone thinks that 100% on-task, uncritical, 0 conflict communication is essential or even normal for work.
It reminds me of an article that claimed Britain has become a more civil and less honest place in the last few decades. (The Economist, maybe?) Perhaps some workplaces are like that too. Whether it's a good thing or not, who knows. People clearly have their preference one way or the other.
This specific case is tied to complicated American politics, but I guess small versions of the same debate are happening in workplaces all over the world. Though maybe, ironically, not in the most intensely policed workplaces.
I'm curious about this. When I've hired, I've always wondered how I can actually tell (a) what soft skills are required for the role and (b) whether a candidate has them.
People sometimes think that's a silly thing to ponder: it's obviously obvious! But at most places I've worked, we spend lots of time defining the technical skills required for a job and handwave the rest.
I guess people assume "they'll know it when they see it". But there's a lot of ambiguity. Parent comment suggests that being comfortable sitting in a conference room for an hour is an important part of their job. In some workplaces that would be an odd requirement. I've worked at places where the important thing was being able to go away and make progress on something for a few weeks.
I suspect there are people with autism reading these threads and feeling disheartened. It would be easy to leave with the impression that neurotypical people expect you to make all the effort and they won't try to meet you half way. Some workplaces are like that. But in all the talk about neurotypical vs neurodivergent, it's easy to forget that neurotypical people are a varied lot, just like neurodivergent people. Workplaces are a varied lot too.
People invented and discovered remarkable things before modern statistics and RCTs (work on puerperal fever, antisepsis and germ theory stand out for me). Humans can make surprising progress with only small scale experimentation and observation.
I prefer the results of large studies. I think modern methods are better than methods from the Scientific Revolution. But people can't always afford to wait a decade or two for solidly replicated results.
The grandparent article claims that Page and Brin were paid by NSF, working on DLI projects while researching PageRank and that the equipment for their prototype crawler was partly paid for by DLI.
If that's true, I'd say they're very fortunate that the Digital Library Initiative existed and that they could put their research into the public domain to reuse it for free at Google. In another context, I'd call the DLI an angel investor and they'd have wanted a slice of that Google pie.
It's not about Cloudflare, it's about the millions of people in Spain who couldn't access a plethora of legitimate, unrelated websites and services because of the block.
I happen to agree that La Liga wildly overreaching is on brand. But I think this is partly about Cloudflare.
What's happening is a reminder of how centralised the internet is becoming. If blocking Cloudflare IPs brings down big chunks of the internet for Spain, that's a problem. Cloudflare could go down for a while, or collapse permanently, or get compromised.
Putting aside my opinions on La Liga overreach, it will also be a problem if companies get to say to courts "Oh, well, if you block those IPs the internet goes down for your country, so let us know what you want to block and maybe we'll get around to it."
Cloudflare might get a resolution from the court that suits them in the short-term. But drawing this to government attention might not suit them in the long run.
How the current US administration is treating European and North American countries makes me wonder if they are serious about taking on China.
As a Brit, I've been surprised by how much my view has shifted in the last few weeks. I used to assume we'd be allies with the US and have a probably competitive and maybe adversarial relationship with China. Now I see the US administration basically saying "we're going to make you pay through the nose for everything" (e.g. taking Ukrainian minerals). So I've started thinking "Well, if the US and China will both behave like that, surely we're best off playing them against each other and seeing who'll offer us more?"
That seems like a very bad deal for the US. So I figure that (a) the administration isn't serious about taking on China, or (b) assumes European and North American countries will roll over, or (c) they think they can go it alone.
Huh, looks like he changed the image in 2024 (based on url). Before today, I'd have said the jwz image was goatse.
Dugatkin, "Winner and Loser Effects and the Structure of Dominance Hierarchies" was an interesting read this morning. Thanks for the reference.
The main thing I take away from the paper is that (in computer simulations) bigger winner effects and bigger loser effects lead to strongly defined hierarchies. Fewer subjects have ambiguous positions.
I'm not sure the paper speaks much to which strategies (e.g. violence vs cooperation) are more common or effective.
A Bayesian prior of 50/50 seems high to me. It assumes that 50% of new disease variants come from lab leaks.
In the last few decades there have been 1-2 confirmed lab leaks per year. And they're often thing like "we found a vial of smallpox we didn't know we had" not new diseases.
Nature very capably produced colds, flus, a bunch of nasty diarrhoeal diseases, the many and varied sexually transmitted diseases, the hemorrhagic fevers, and so on. For "some new disease variant that I don't know anything about", my prior would be more like 1/99 lab leak to natural origin.
I can't immediately think of a major US naval deployment in the North Sea outside the big NATO exercises every 1-2 years. Were you thinking of a specific shipping lane?
Like if this is a crowdsourcing project, why not do a first pass with an LLM and present users with both the image and the best-effort LLM pass?
Possibly for the reason that came up in your other post: you mentioned that you spot checked the result.
Back when I was in historical research, and occasionally involved in transcription projects, the standard was 2-3 independent transcriptions per document.
Maybe the National Archive will pass documents to an LLM and use the output as 1 of their 2-3 transcriptions. It could reduce how many duplicate transcriptions are done by humans. But I'll be surprised if they jump to accepting spot checked LLM output anytime soon.
This is great advice and I've seen it work for a few people. (Not specifically for cybersecurity.)
The short version is when people can ease sideways to a completely new role at their current employer. The longer version involves getting a new job doing the same thing, but at a company that does a lot more of the thing you want to do. Initially you do the thing you were doing before, but there's more opportunity to shift sideways when you get a good rep.
I'll confess I was sceptical about this but, at minimum, the database seems to exist.
There's a company called Carfax that I'd never heard of. Their EU site seems to provide basic reports about the VIN, whether the car has been written off, etc. Those basic "Is this car sale a scam?" checks are common in the UK.
But the site also makes a big deal about "Get the American report!" So I googled "Carfax oil change" and found people talking about the oil change history in the reports [0]
In the UK it was traditionally common to keep a car log book where you recorded all maintenance and might get the garage to put their stamp on it, to prove to a future buyer that you'd looked after the car. But having a garage enter that info into some random company's database, and maybe not telling me, would be disappointing.
[0] https://www.toyotanation.com/threads/oil-change-history-when...
Are you suggesting that a process once done using pen and paper can't possibly be complicated?
I have no insight into the DVLA, but the idea that no paper process could ever be complicated is really funny. The UK enjoyed/loathed centuries of bureaucracy before computers were invented. At one point getting a divorce required an Act of Parliament specifically naming the unhappy couple! Being restricted to pen and paper hardly inhibited the human ability to create complex systems.