It’s not the ‘Linux on’ subsystem, it’s the ‘Linux on Windows’ subsystem, so it’d have to be Linux on Windows Windows Subsystem, which would be silly. It can’t have a colon in it as some command-line tools take a subsystem as an argument, and traditionally, Windows command-line tools have used colons the same way Unix has used equals, i.e. to separate an argument name from its value, and parsing that gets harder when you’re expecting colons in the value, too.
Windows has subsystems. They’re called Windows Subsystems. This one’s for Linux. However you slice it, the initialism has to have WS in it.
There’s a pretty good reason to think it’s not going to improve much. The size of models and amount of compute and training data required to create them is increasing much faster than their performance is increasing, and they’re already putting serious strain on the world’s ability to build and power computers, and the world’s ability to get human-written text into training sets (hence why so many sites are having to deploy things like Anubis to keep themselves functioning). The levers AI companies have access to are already pulled as far as they can go, and so the slowing of improvement can only increase, and the returns can only diminish faster.
If LLMs aren’t going to reach a point where they outperform a junior developer who needs too much micromanaging to be a net gain to productivity, then AI’s not going to be a net gain to productivity, and the only productive way to use it is to fight its adoption, much like the only way to productively use keyboards that had a bunch of the letters missing would be to refuse to use them. It’s not worth worrying about obsolescence until such a time as there’s some evidence that they’re likely to be better, just like how it wasn’t worth worrying about obsolescence yet when neural nets were being worked on in the 80s.
Usually, having to wrangle a junior developer takes a senior more time than doing the junior’s job themselves. The problem grows the more juniors they’re responsible for, so having LLMs stimulate a fleet of junior developers will be a massive time sink and not faster than doing everything themselves. With real juniors, though, this can still be worthwhile, as eventually they’ll learn, and then require much less supervision and become a net positive. LLMs do not learn once they’re deployed, though, so the only way they get better is if a cleverer model is created that can stimulate a mid-level developer, and so far, the diminishing returns of progressively larger and larger models makes it seem pretty likely that something based on LLMs won’t be enough.
To be fair, if I had all that money, I’d probably just pay someone to figure out how to make it do the most good, and continue spending at least some of my time shitposting. It’s okay to have hobbies, but it’s bad to hoard the money or invest it in evil.
I’ve found this is really dependent on placement. If I put my libre a couple of centimeters away from the region I usually use, it’ll read low all night, but as long as I stick to the zone I’ve determined to be fine, it’ll agree with a blood test even if I’ve had pressure on it for ages. Also, the 3 is more forgiving than the 1 or 2 because it’s smaller than the older models, so affects how much the skin bends and squishes less.
Plenty of TVs are capable of radioing your neighbour’s TV and piggybacking off their internet connection, so if it’s not in a Faraday cage, it might be overconfident to say it’s never been connected to a network.
It wouldn’t be Rocko’s Baselisk if it didn’t do things that hurt.
What kind of time frame were they testing over? Not seeing any significant burn in means something completely different if they’re testing for one year versus ten, especially for people who don’t like replacing things that aren’t totally unusable yet.
Some of the charity is self-serving, e.g. eradicating diseases means he’s less likely to catch them (and really any billionaire not funnelling funds to pandemic prevention etc. is being moronic), and founding charter schools on land he owns so over the life of the school they pay more in rent for the lease than they cost to build is just a tax dodge. Most billionaires are just so evil that they won’t spend money on themselves if other people who aren’t paying also benefit, so in comparison, Gates’ better ability to judge what’s in his interests makes him look good.
Investors have been happy to incentivise companies to hire idiot CEOs and managers who say the right buzzwords but reduce output by making bad decisions and only hiring people who don’t think they’re bad decisions, so an automated buzzword-dispensing idiot isn’t necessarily going to seem to investors like a downgrade compared to what they think most workers are. They’re just as likely to think AI lets them invest in companies where even the lowest tier employees are potential CEO material, and continue not noticing that the per-employee efficiency keeps going down. Data showing that layoffs nearly never pay for themselves doesn’t stop stock prices soaring whenever one happens, so I wouldn’t expect data showing AI makes companies less profitable to stop stock prices going up when a company announces a new dumb way they’ll use it.
There was quite a lag between the variable-rate mortgage rates going up and everything noticeably exploding, so lots of people who were aware there was a real risk of things going tits up decided that it hadn’t and therefore wasn’t going to and had stopped looking for signs by the time they started to appear.
The discussion on the talk page was basically nuh uh, loads of reputable sources like the Israeli and German governments say there’s no evidence of genocide, and even if they’re biased, anti-genocide NGOs are more biased because they have to accuse nations of genocide to justify their existence, with people responding to point out that’s not how Wikipedia’s rules work, and if it were, they’d have to rename the pages on various other genocides because there are very few that no nations deny.
In principle, copyright law doesn’t stop there being a system that lets you redeem the same key from Steam, Epic and Gog as long as it’s the same person behind all three acounts. There’s already a degree of precedent for this - when a publisher generates Steam keys to sell at other retailers (whether they’re codes-in-a-box at a physical shop or an online retailer like Humble Bundle), they don’t have to pay Valve a fee, but the keys can be redeemed on Steam and work just like if you’d bought the game from the Steam store where Valve would take a 30% cut. Valve probably don’t think it’s in their interest to make libraries transferrable/sharable between Steam and not-Steam, but if they change their mind, and the competitors that they’re building the transfer/sharing system with also thought it was in their interests (which is unlikely to happen at the same time), there’s nothing stopping them building it.
If you trust Adam Smith, an efficient market would quickly make new goods and services available for the lowest possible price. People are supposed to be incentivised to sell things, and competition is supposed to stop an excessive amount of money being diverted to profits instead of reinvestment and price cuts. There are supposed to be systems in place that ensure that there are always opportunities for competition, but the wealthiest have the most power to erode those mechanisms and the most interest in eroding them, and neoliberals think an Ayn Rand novel is the definitive text on how capitalism works rather than Adam Smith’s work, so even if you land in the capitalism makes everyone better off mode, it doesn’t last for long before it falls back into fuedalism with extra steps mode.
It’s pretty obviously both things that were a problem. A mediocre prebuilt with a $200 premium over other mediocre prebuilts was unattractive, but plenty of people do buy overpriced mediocre prebuilts. The killer feature being that nearly no games were compatible was going to kill it even if the price was right, though.
Now we’re in the era of Proton, and games just work, if the price is fair, it’ll sell. If the price is you cannot get anything like this without spending much more like the Deck or how consoles used to be priced when the hardware made a loss and the profits came from games, then it’ll sell really well.
Plenty of bank apps work just fine. None of the ones I’ve tried had problems, except Santander, which works perfectly after changing a setting.
The scary dystopian part is the ability to work out that the account belonged to someone who hadn’t used it for a decade rather than just that they could see what had been posted. The Internet Archive doesn’t let you ask it what someone’s Digg username was.
It’s not so much about time (although I have played a couple of things that would take a ridiculously long time to save or load), it’s about the number of chances to make a mistake. If you only save ten values, it’s really easy for a programmer to verify that they’ve got everything right, but if they save ten million, there are a million times more opportunities for mistakes to sneak in and it’s much harder to notice each mistake, let alone fix it.
Fallout 4 is a bit of an odd duck here as the save format for the BGS games is basically just another ESM file, so reuses all the same serialisation and deserialisation mechanisms. Most games don’t have multiple places the game data can come from and a way to combine them as they’ve not got an engine designed with this kind of modding in mind, so there’s nothing to reuse in this way for saves. Given the general standards of engineering from that studio, if they didn’t have this as a core feature of their custom engine for nearly three decades, and instead had to come up with something from scratch, they’d absolutely mess it up or have to simplify the saving system.