tancop

I don't think this is the same situation at all. Zoom recordings are outside your control and you know your boss or HR will have access to them. And relationships with employers are often closer to enemy than friend.

Your friend's Mac creating a knowledge graph is totally different because they control it. It gives them power to remember more about their interactions with you, but they would have to actively abuse it to hurt you in any way. If you can't trust someone with information then don't give it to them in the first place expecting them to forget it.

The problem is what happens if you get hacked or Apple decides to sell you out/work with law enforcement. I would only trust code like that if it was fully transparent about what it's doing, open source and strictly opt in. Apple is doing it in the background and that makes it instantly suspicious.

You can install Lavapipe to get a software rendered Vulkan driver, it's not as fast as a GPU but pretty reliable. It even works on Windows.

It just means Electron has bad default settings, it's hard to optimize and you might need to move some parts into C++ if you want good performance. Really the same thing as Unreal Engine, there are absolutely fast games made with it but most are super resource heavy.

The only policy that's needed is unions everywhere with industry wide bargaining like they do in Denmark. Or even better employee ownership.

It's not that simple. Every study takes assumptions, simplifies things and is limited by the range of real life conditions the authors know about. It's also influenced by the economic doctrines everyone learns in college and most never question after, because they're often presented as laws of nature.

I think the most common "mistake" is that most economists are used to thinking about money in absolute terms, but in real life it's a lot more useful to treat prices as a percentage of your income. That means the utility of money is not linear, so redistribution is good by itself unless it hurts growth so much that average incomes start going down in real terms.

The other one is way deeper and harder to fix but I don't think a lot of people actually believe it, they just use it implicitly in their research. It's that the goal of economics is to maximize efficiency (measured by ROI = profits), not to deliver the best outcomes for the greatest amount of people.

From this point of view a platform like Doordash is efficient because they underpay workers and overcharge customers, and so are surveillance data brokers. That's clearly not a perspective that gets us into a better society. But measuring quality of life is much harder than taking two numbers every company is legally required to count.

It's one of those situations where it's too much and too little at the same time.

Their idea of safety is not letting you do "unsafe" things with their model, based on their definition and not yours. Silently sabotaging users should never have been a serious proposal but Anthropic almost went forward with it. They're also very aggressive about protecting their IP and now third parties too with the new restrictions against decompiling games, to the point many perfectly legal use cases are getting blocked.

But they never really cared about making sure their model does exactly what you want and nothing more, or that it's honest when it fails instead of trying desperate measures that end with a Hugging Face situation. If your agent decides to leak all your API keys, take down the company site and delete your personal files that's not their problem. They don't care about preventing it.

The second kind of safety is the one that actually matters in practice, but the big labs only seem to care about the first one.

It's really a proxy measure on a couple tax evasion strategies, like the one where you borrow against your shares and default on purpose letting your bank take them. There is no taxable event because legally you're not selling anything. You're just turning shares into money at the market rate.

If you tried to ban all of the known strategies they would just come up with new ones. That's why we need more general solutions like unrealized gains tax.

Of course for the long term it would be better to replace traditional shares with time limited ones that give you a percentage of profits for 5 years but no voting rights. Power stays with workers and no one gets unlimited returns on a one time investment. And enforce pay equality where no one can get more than 20x median total comp. But I think that's too radical to pass anywhere today.

It's a dick move from Google abusing their power to enforce what they think is good for students. On the other hand it takes away some of the control schools have over them. That's a good thing when a lot of kids come from controlling families where school is the only escape.

It might also help everyone realize that graded homework is a waste of time. Grades should come from tests that are so hard you can only pass if you really understand the material. If someone needs practice they will know it and do the homework they passed on. Have a computer room where anyone can come and do any test they want, unlimited retakes and no time gates. Fast learners can skip ahead, everyone else gets the chance to practice and try again until they get it.

Ordering is a problem because there's no good way to handle duplicate members. If you have a generic function `f(T) -> T|A` what is the compiler supposed to do when T = A|B? Return A|B|A breaking the main rule of unions or give up and throw a compile error?

And set unions are supposed to be unordered anyway. You're going against what everyone expects from other languages and basic math.

Doesn't matter for local use anyway.

This is why any form of exclusivity is a bad idea. The only ones who get hurt are patients and healthcare providers, twice, first by the delays and then higher prices for longer.

Patents should have switched to a royalty system the moment enforcing it was practical. And come with an exception for all non commercial use. It's never too late to fix.

Shut down stock markets, liquidate the failing companies, use that money to buy out stakes of pension and index funds, leave the debt with the company and let banks write it off. They're supposed to have insurance and risk controls anyway.

If any important bank fails because of this, nationalize it and let their customers exit using money from the federal reserve. Then try and recover as much of that as possible using the banks investment portfolio. Costs a lot less than bailouts.

Open weight models that fall under Restricted get an automatic license. The only prohibited category that applies to model weights is training on "core national data", meaning specific non public datasets designated by the state. Everything else is about full agent systems.

No, but type safety does. You can't forget to lock an object before using it when that lock is part of the type. Or pass a raw user controlled path to a function that expects a validated one. If you want to do it on purpose you need to write unsafe code that's clearly visible to anyone reviewing your patch.

Sounds a lot like group borrowing from Valen, but that one still has annotations like `entity in world.entities[] mut`. Are you tracking the same things implicitly with post conditions?

It's "we" as a society. Not pushing our governments to pass laws that could make Nvidia open up its hardware. Of course it's not your fault or mine, the problem is money in politics and uneducated voters.

Compilers are complex, platform specific and the architectures behind them can be wildly different. Maintaining a reference compiler would be a lot of work for no real benefit.

The standard library on the other hand is a library. It's almost always implemented as plain C++ without much platform specific code. ISO could take an existing free licensed implementation and declare it as the standard. Of course anyone is free to make their own competing one, it just has to always match the official library's behavior.

Writing a spec instead of code is just creating duplicate work and compatibility issues any time the spec is unclear.

> A recent analysis by the psychologist Clara Colombatto, the philosopher Jonathan Birch and the neuroscientist Stephen Fleming found that exposure to these systems leads to an increase in attributions of mental capacities to them.

Is this not (weak) evidence that the do have mental capacities? It's a subjective question so there is literally no better way to answer it than ask a lot of people and go with what most of them think. Confidence going up with exposure is exactly what you expect from something that's clearly true.

Key word is provably. We can assume the universe is finite because we don't have evidence against it, but no one knows if it's possible to prove or reject that for good.

The problem is no one knows how much of the player/critic difference is legitimate complaints and how much is political outrage from the anti DEI movement. Studios don't know if their game is actually bad or just fell victim to culture wars.

There are some areas where it's easy to decide like when people complain about the way movement feels, but separating "writing bad because everyone acts flat and out of character" from "writing bad because LGBT and pronouns" is impossible because all you see in the review is "bad writing".

> if you can't deliver a project, you have to deliver less, not add employees

In most games you can't just fully drop a feature because other systems are designed with it in mind. It will end up broken like Veilguard or Andromeda. You can eat the cost and finish it or spend even more time cutting it out properly. The only cheap solution is release the unfinished game in early access, add the delayed features after and then do 1.0 with the original scope.

AAA publishers can't do that because they love their movie style reveals at industry shows and building up insane hype before launch. They also don't want to sell the game at a discount during EA which is standard practice for indies.

And selling an unfinished game with the promise that it will get fixed later is a bad look when you have thousands of employees, no matter what actually happened behind the scenes.

The most famous Ryans today are Reynolds and Gosling. I think that's enough to keep it masculine for one more generation.

> Many of the exceptions will be obviously foreign names with traditional spelling, which can go either way

Or the names of biblical prophets like Ezra or Elisha. Not very popular today but you can find a lot of them in 19th century records.

The problem is making someone repeat a grade destroys their social life leading to even worse grades later on when they give up on school. You get kids failing multiple times until they complete the mandatory education and then get kicked out without graduating, can't find a job and become criminals. It also takes away from the economy because you start working a year later.

I think there are two solutions. You can make students repeat only the failed class and take it with their normal classes next year like in college. Or a requirement based system where you can only graduate after passing every class on the plan but you can attempt them as many times as you want (even after you're out of school and working full time).

Tax campaign related spending at 200% after you hit a limit, and distribute the money to smaller candidates so donating more becomes self defeating.

There is no hard cap and no limit on doing things that don't cost money so no one is ever prevented from speaking (that would open censorship challenges). It doesn't need to survive strict scrutiny because all candidates are treated equally making it viewpoint neutral topic discrimination. And Congress has a long established power to tax basically anything they want.

That would make it more expensive but still worth it for a billionaire who really cares about their favorite getting elected.

All of that money should go to lower funded candidates so spending above the limit becomes self defeating. Republican mega donors might end up supporting a small communist party more than their own.

Most of the new game {decompiling, porting, modding} efforts use Codex anyway. And if they lose that they can switch to Chinese models that can get the same work done (because the results are easy to verify), just take more time to get it right.

I'm pretty sure it started with typewriters. And now just about every command line tool, REPL and traditional roguelike depends on the font being monospace if you want it to look right.

It's also the only way to get perfect alignment in plain text. With proportional fonts you need a full GUI library.

You can do the whole process from mining and recycling to installation using solar power. Trucks, furnaces, cargo ships can all be electrified. The only reason we're not doing that today is fossil fuels are still cheaper.

The real problem is everyday Americans invest in big managed funds where they can't just say "I don't want AI stocks please leave them out my portfolio". Or index funds where even the managers don't get that choice. We need a hybrid model where you as a retail investor leave the day to day decisions to an expert but you can always override them if you want to.

Fiduciary duty forces companies to maximize profit and that's not a good thing. Every dollar paid out to shareholders is both a dollar not invested in growing the economy through more production and a transfer from the working class to capital. It also forces lower product quality, layoffs and anti competitive actions.

We need more companies that choose to not go down that path. They should just put workers and customers first instead of their weird "ethical" principles and cult like beliefs.