By minutes watched? Futurama. I've re-watched it like half a dozen times now. 

If and only if they give it up-to-date date, and ideally if they give it internet access, I thing $5-10 would be okay. Assuming it really always works, is fast and responsive, and has no sort of advertising of any kind.

Congratulations on the hot take. Great job.

Not planning ahead is not planning ahead.

Sure, but 120 degree “hot” water is really hot compared to a frozen window. While merely warm water, or room temperature water, is not nearly so big a problem.

WARM not hot. But also... you could just use a de-icer.

I’ve been using a free program called “Input Director” which works very well, and supports far more configurations than this. I’m sure this is probably slightly snappier and can’t be messed up by having a slow network connection. But the flexibility of a software solution kinda makes hardware KVMs lackluster.

Okay, let’s give this a try. Are we all willing to be open minded for a moment? Yes? Okay, here we go...

Not really. ChatGPT specifically can’t do any of those things, but can be “tricked” into pretending that it’s doing those things. It can generate answers that look and sound like it went online, or that it ran a VM. But in reality it’s just using its LLM to produce output. This is similar to showing a human that is…

At the moment ChatGPTs greatest weakness is its limitation to only give you information based on the training set it was given. If it, or another LLM, were able to instead access the wider internet through a search-engine like system it could located, digest, analyze, and summarize information without needing to…

There’s definitely a fine line to walk here. On one hand, if you go to see a movie because you love the actor who was in the trailer, and they’re not even in the movie you are rightfully unhappy. If you go see a movie that appears to be a comedy in the trailer, and it turns out to be an action horror movie... you…

Except no, that’s precisely what it isn’t. That’s what makes AI (machine learning, to be more accurate) what it is. It LEARNS what art looks like, and then makes things that look like art. This is exactly what brain does.

Not really doxing when the information is publicly available flight information. If Elon doesn’t want people to know where he’s going, he can fly commercial or drive his Tesla. One of the costs of the privelege of having a private jet, is that the information about its movements are public information.

Wait, there's more context to this story? History??? Too bad Gizmodo writers don't bother doing any research. This comment was thr most interesting part of the story! 

The gyms actually have an order? I thought they scaled, I took them in a pretty much random order. I guess I may have gotten lucky.

The funny thing, is that I said Larry also, but the real reason was that I was answering which gym CHALLENGE was the most difficult, which was the stupid riddles and food ordering crap. Having to walk around town and find the clues was such a pain in the ass.

I’ll just put this out there, as I’ve seen it posted in a few places:

I think not using "sub-race" for humans is maybe a good idea though. But it's fair to say using it for others is also not great. 

They actually did acknowledge more half-species by dropping the existing halfs, and adding a new rule for mixed characters. It's not exactly the best implementation IMO, but they did make some changes. 

Eh, but in fantasy the term “race” has been in use for a very long time as “a type of thinking being.” I don’t think it was ever really a source of confusion. Plus, in the real world "race" isn't even a biologically meaningful term (turns out that genetically "race" doesn't actually exist).