Guys, you can laugh at a joke. The AI doesn't win just because someone upvoted a meme. Maintainability of codebases has been a joke for longer than LLMs have been around because there's a lot of truth to it.

Even the most well intentioned design has weaknesses that we didn't see coming. Some of its abstractions are wrong. There are changes to the requirements and feature set that they didn't anticipate. They over engineered other parts that make them more difficult to navigate for no maintainability gain. That's ok. Perfectly maintainable code requires us to be psychics and none of us are.

There's historical truth to this. In toothpaste, no less.

Ingredient: Asbestos

Comes from: naturally occurring mineral

Used for: mild abrasive

Only Murders in the Building

How does that make sense lol. If everyone is 5 minutes late, then the whole schedule shifts back 5 minutes. The only way it builds into an hour long wait is if they're being overbooked on an over optimistic schedule. Hoping they'll only take X minutes but they actually take X+5, for example.

I think the point was less about making the messages undetectable as it was about making them unjammable. In order to stop their transmission, we would have to essentially shut down GPS for the entire EU. So you might use that frequency to send critical, must-have messages.

Yeah I was trying to figure out this as well. It's about gravitational time dilation. The increased gravity at the center of the sun means that about 25,000 fewer years have passed there than at the surface, since the birth of the sun.

There's definitely a yo momma joke in there somewhere

At a minimum, the agent should be compiling the code and running tests before handing things back to you. "It references non-existent APIs" isn't a modern problem.

Um. Lifting the sanctions lets tankers carrying Iranian oil sell it on the global market and then turn around and buy more from Iran. How does that not make Iran money?

It's not the storage of the information that matters as much as the presentation. Google's search index stores a huge amount of copyrighted material, even losslessly. But they only present small snippets at a time which is not considered copyright infringement. The question really is whether or not the information being presented by the models is in a format which is considered copyright infringement. So far, courts have not found that they are.

I think 100% autonomous robotics and driving is still at least 5-10 years away even with large research teams working on it. I mean truly robust AI which is able to handle any situation you could throw at it with zero intervention needed.

A few do. You draw in cold water from the depth of a lake or fjord and pass that through a heat exchanger before dumping it back in the reservoir. It's more common in power plants. There's only so many places where the geography works out for this though.

Funnily enough, if you actually look into the source of these data center water consumption memes, they typically count that circulated water as "consumed" by the data center despite the fact that, you know, it doesn't actually go anywhere.

String theory has made predictions that have turned into real world results. For example, it accurately predicted the properties of quark-gluon plasma. Sure, quantum chromodynamics can be used to make these predictions too but it's vastly more difficult.

That's why scientists keep working on it, because the value in a model is in its predictive power and a model's ability to make predictions elegantly is incredibly important for doing real world work.

The truth of the underlying perspective a model has on the universe is kind of secondary. Newtonian mechanics was wrong. Einstein's relativity was wrong. Quantum mechanics is, at the very least, incomplete. String theory most likely is wrong too. But they're still useful and valuable in the right contexts.

I don't think being falsifiable is the most important part of a model. It's their ability to make accurate predictions about what happens in reality. All models are wrong. Some are useful, the saying goes.

Newtonian mechanics, which we know is wrong, can still be used accurately for a lot of near-earth orbital calculations. It got the Apollo missions to the moon.

It's my understanding that one of the most useful things about string theory is that it can make some really complicated or intractable problems much easier to solve within its frameworks. That's valuable, regardless of how truthful it is that everything is made of 11 dimensional strings.