1) It's true that AI algorithms are used extensively in building search results, and have been for a very long time.

2) What schools are banning is generative AI, which is fundamentally different from most AI used by Search.

This is such a blatant knee-jerk reaction. Google is throwing its monopoly power around to try and force AI adoption. It's disgusting.

Now it's time to wait on tenterhooks for the next few couple of months while Republicans cry "rigged election" and try to use that as an excuse to overturn the results. The courts have so far denied their attempts at every attempt. Will our court system hold up a little longer? Will Republicans succeed in getting some districts to avoid certifying election results?

This is called price fixing. And it is precisely the kind of thing that we have anti-trust laws for.

But will the DOJ or FTC bother to look into it? I have my doubts.

I've been working as a software engineer for over ten years now. If you can learn to code decently in one language, you can pick up a second or a third pretty easily. In my career, I've coded in C, C++, Java, JavaScript, C#, Go, a number of dialects of SQL, Python, and others. Employers really should not care much at all which languages a person knows. The question should be how well they can code, not the language.

There's a very easy way for this to be an "oopsie". As in, it was a change that was generated and discussed, but the final decision to make the change was never made, then somebody clicked the wrong button and released the wrong version.

Which makes me sad. Misinformation is a massive, massive problem online, and efforts to curb scamming and other forms of misinformation through payments could help.

This is exactly what I would expect. Most households who aren't near poverty can find ways to cut costs on food pretty easily, and by much greater monthly amounts than streaming services cost. I imagine most households will spend as much on food in days what they spend on streaming services for a whole month. Why wouldn't food be a more obvious place to cut costs?

Depends upon the degree. If they're expecting reduced demand, then it only makes sense to reduce supply to compensate. Otherwise there will just be GPUs sitting on shelves. But, of course, if they go so far that it becomes impossible to find the new GPUs after release, then yes, your criticism will be accurate. I suppose we'll see how it shakes out in the end.

In a way, yes. All galaxies beyond a few billion light years away (I don't recall the exact distance), which is most of the visible galaxies, are now and always have been receding at faster than the speed of light.

How? Well, the speed of light limitation gets a bit more complicated in curved space-time. Once you have to take curvature into account, which is necessary at large distances, the speed of light rule gets weird. The real rule only works at the same time and place: nothing can outrun a light beam. There is no notion of a speed-of-light limitation for far-away things because the concept of relative speed becomes very fuzzy in curved space-time.

Think, for example, of two cars traveling somewhere on Earth at the exact opposite ends of the Earth. Both are traveling North. If you simply looked at their three-dimensional velocity vectors and subtracted them, you'd conclude that their relative speed is zero: they both have the same velocity in the same direction. But if you think about the fact that the Earth is curved, the two cars are getting closer together relative to the Earth's surface. So if they're both moving at 25mph, then the shortest distance between them across the Earth's surface is dropping at 50 miles per hour.

This is why space-time curvature causes no speed-of-light limitation for far-away things: there's no one way to define how fast one thing is moving relative to another. There are multiple ways. Same is true for distance. Are the two cars 7900 miles apart (the distance through the Earth) or 12,450 miles apart (the distance across the Earth's surface)?

So whenever you hear a distance to a far-away star (27.6 billion light years here), you have to understand that there were some arbitrary choices made on deciding what we even mean by "distance". Astrophysicists have something like five different commonly-used definitions of distance, all with different definitions and all which will give different values for "distance". They all will give the same answer for things that are close-by, mind. But the answers get very, very different when things are far away. Because curvature.

So, curvature is weird and distance and velocity are complex, but what does the 28 billion light year number actually mean then? Basically it means that the light was emitted from roughly 3.8 billion light years away. But the universe has expanded by a factor of roughly 7.2 since then, so the current distance is 7.2 * 3.8 billion = 28 billion light years away (roughly).

They get all of these numbers from the star's redshift. From other data sources we know how fast the universe has expanded over time, so we can calculate how far away things are (given a definition of what we mean by "how far").

It's highly unlikely that they're hosted on IPFS in this specific instance, as it's a promotional idea from a company selling shoes. The NFTs are probably just links to images stored on AWS. It'd be trivial to produce a court order for the destruction of the data in the underlying link.

Last I checked, EVs needed to spend about 1/3rd of the travel time charging for long distances. EVs are great for shorter distances where you can charge at home or leave it to charge at a station while you do other things. They're not so great when you want to, say, visit a friend in another state and don't have a way to charge at the destination. This would neatly solve that problem if it were implemented in enough places. It would probably also make electric-powered public transportation more feasible in more places.

That's not how I read it. This isn't the time spent actively updating. Rather, if the machine isn't online for 8 hours at a stretch, the auto-update feature running in the background may simply fail to complete and the actual install will never start.

That's a remarkably bad argument. Regulation isn't a case where you can easily point at one thing and look at its effects. The results of regulation are the sum total of a variety of regulations working together, often in complex ways.

The fact is that what we really want is an efficient system. There are two general ways that society has found to make efficient systems: 1) Use light regulation which encourages competition, 2) Enforce a monopoly which is tightly-regulated. If you want the worst of both worlds, you enforce a monopoly but use light regulation. This guarantees bad outcomes, because all of the benefits of the monopoly (e.g., limited number of wires being run everywhere) are completely outdone by the lack of incentives of the monopoly to do anything that resembles economic efficiency. And it's the reality of most internet service in the US.

Properly implementing regulations is a difficult business, to be sure. It means that the regulations need to be nimble in response to data that unintended consequences have occurred. This is just a fact of life: even if you want to go for the highly-competitive market, regulations will be required to keep it highly-competitive. Competitive markets are fundamentally unstable because the biggest competitor in the space will always find ways to throw its weight around until it becomes the only competitor in that space unless it is counterbalanced.

In practice, these kinds of banal arguments against regulation do nothing but benefit monopoly power.

Exactly this. Most local governments have mandated monopolies for ISPs. Until that changes, the only way we can get better service is much stricter regulation.

There are no heroes here. Only villains.

It makes very good sense that Epic should pay some fees to make money off of an app on Apple's platform. Their attempt to pull their payments off of the platform in order to avoid the fees is pretty flagrant cheating.

But the 30% standard fees (both iOS and Android) are exorbitant. There is simply no good reason for fees that high. I suspect it's a relic from when the amount of money to be earned on these services was far smaller, so that Apple and Google really did need higher fees in order to justify their investment. But that just isn't the case any longer. These ecosystems are huge. Keeping the fees at the same level is today purely a result of their monopoly power combined with collusion (the two companies should have to compete with one another by lowering such fees to attract more apps to their respective platforms, but they are both deliberately avoiding this).

If we had a better government in the US right now, there might be an anti-trust investigation to be had here.

Not entirely. But mostly.

A hack like this almost certainly changes little to nothing about how the Tesla-released code behaves. It likely just sets a few flags, which in itself should have no impact on safety whatsoever (no moreso than Tesla setting them).

But there is always a chance that the hackers either made a mistake or slipped in some risky code. I expect the chance of that is very low, however. Tesla can certainly verify this. But instead they'll rely upon this theoretical risk to justify forcing consumers to spend more money.