astronaut_sloth

I study AI, and have developed plenty of software. LLMs are great for using unfamiliar libraries (with the docs open to validate), getting outlines of projects, and bouncing ideas for strategies. They aren't detail oriented enough to write full applications or complicated scripts. In general, I like to think of an LLM as a junior developer to my senior developer. I will give it small, atomized tasks, and I'll give its output a once over to check it with an eye to the details of implementation. It's nice to get the boilerplate out of the way quickly.

Don't get me wrong, LLMs are a huge advancement and unbelievably awesome for what they are. I think that they are one of the most important AI breakthroughs in the past five to ten years. But the AI hype train is misusing them, not understanding their capabilities and limitations, and casting their own wishes and desires onto a pile of linear algebra. Too often a tool (which is one of many) is being conflated with the one and only solution--a silver bullet--and it's not.

This leads to my biggest fear for the AI field of Computer Science: reality won't live up to the hype. When this inevitably happens, companies, CEOs, and normal people will sour on the entire field (which is already happening to some extent among workers). Even good uses of LLMs and other AI/ML use cases will be stopped and real academic research drying up.

Just from a quick look at https://fediverse.observer/, it looks like the Fediverse is mostly steady at 1-1.25 million monthly users (give or take) over the past two years with a slight decreasing trend. I think there are some reasons for this that are not entirely in our control.

There seems to be a global sentiment of disconnecting from social media and the internet in general. So, I wouldn't be surprised if ever platform is seeing a decaying user base. Anecdotally, among the people I see in real life, there is a general sense of exhaustion with online spaces. Whether it's from corporate-own, enshittified platforms to even places on the Fediverse, the people with whom I interact tend to find the entire thing hollow. They've trimmed down to one or two platforms (if that). In fact, I've even started to get that way. In the past, if someone were wrong and arguing against a point I made, I'd engage, especially if it's in something that I have expertise. Now, why bother? There's no use arguing; people have little interest in admitting fault or engaging in good faith (again anecdotally). That said, I'll concede that the Fediverse is a bit better on that front, but not by much.

Then there's the alternative nature of the Fediverse. It's been rehashed over and over about how "difficult" it is to get on and use. It's not actually that hard, but the barrier to entry is an extra step. That small extra step frightens people away from even joining. The only time that barrier gets broken is when a "legacy" social media platform does something anti-user. Then there is a refugee wave that comes in and goes out leading to a modest durable increase in users. Recently, there just hasn't been a major controversy on a major platform that leads people here.

Now, my final thought on this is to ask: Is a small and steady-ish population (despite modest decay) actually bad? In my view, I don't think it is. Being smaller and with a smallish barrier to entry means that we exclude a sizable number of the low-effort population. So, there's less (no zero) slop here. Plus, discussions, when had in good faith, can be much deeper and less filled with stupid low-effort jokes. Overall, I'm not too concerned with the number of people on the Fediverse. Growth isn't necessarily the best thing. Even so, with the way most mainstream platforms are going, it's inevitable that they will do something stupid that drives more people to the Fediverse at least for a time.

TL;DR: The monthly population is mostly steady with a modest decay. Most social media is likely seeing similar trends. I don't think the smaller userbase is that bad of a thing.

Somewhat related but I always half-joked that Clarence Thomas' anti-civil rights rulings were a long con to outlaw interracial marriage so he could divorce Ginny.

I already know the vote will fail, and even if it doesn't the Senate won't convict. But, Jesus, the Democrats need to do something. Even a symbolic gesture is welcome, and it's better if it can gum up the works and make Trump's agenda come to a screeching halt.

this “cameras for everything!” idiocy.

That's why I'm so impressed with how well it's actually working. When they get off that really weird self-imposed restriction, it could be an interesting technology.

Not great performance at all.

That's better than I was expecting to be perfectly honest.

I'm pretty impressed with the technology, but clearly it's not ready for field use.

Great article! For a few years, I was always deterred from projects because they had already been done and better, so there was no reason to do it. Now, though, I just enjoy implementing things in my own janky way and learning a bit along the way.

You do get promotions in the Army reserve.

Correct.

They are taking jobs that could have been a promotion for an experienced reservist.

Probably not. This little think tank was just stood up. Title 10 gives the maximum numbers of officers per service per grade ( https://www.law.cornell.edu/uscode/text/10/523), and I wholly doubt that five new O-5s make an actual difference to getting to that cap. I also doubt we are anywhere near the maximum numbers of authorized officers. There is a retention crisis after all.

Beyond that, they would have to compete within their own competitive categories, so whatever category these people are in is the only one (minorly) affected. In fact, what is most likely is that the Secretary of the Army authorized to add five to the quota for that competitive category to mean that no one gets negatively affected. When officers are directly appointed to senior grades (which does happen), it's not a big deal.

Look, I agree that this is bad, but it's not malicious...just dumb and a waste of time and money while cheapening the service career officers have. In the end, this is just a publicity stunt and an eval bullet for some General somewhere. There's plenty more to be outraged about from this administration than direct commissioning five idiot executives who will likely not do anything of substance.

LTC’s either have command or staff positions and often approve things like operations orders and contracts.

They are not going to have command. The article itself says they're going to some "Innovation Corps," which just sounds like some boondoggle assignment on a staff. They will have no actual Army job. They're just going to be pushing contracts to their companies. That's bad, but it already happens, so it comes out as just nothing. They are almost certainly not going to be in anyone's chain of command.

Plus, I'd bet they'll probably be at the Pentagon. An O-5 there has as much authority as an O-1 in the "real" military.

There’s virtually zero chance they’ve been put there for no reason

Probably just some flag officer's good idea fairy or a way for a flag to secure a job after retirement. Again, not good, but very par for the course for DoD stupidity and/or corruption.

There are plenty of actual things to get outraged over. Having some tech bros play Army as O-5s is not that important. The DoD already gets bent over a barrel by Palantir and other companies to use their software; it's wasteful and supports terrible companies. That's what to get outraged over, not some idiots being appointed as O-5s.

My decade of experience working with the military says otherwise.

There is a small minority who joins to kill people, probably like 5%. It's way too high, but it's still in the vast minority. Most officers and NCOs, in fact, prefer thinking personnel so that war crimes aren't committed and laws are followed while still accomplishing a commander's intent. While I'll admit that there is a cavalier attitude toward collateral damage (which is a separate issue), war crimes, a la Eddie Gallagher, are generally not tolerated.

Of course, with a vet bro SecDef, war criminals will become more numerous and have a sense of impunity. It's why the vast majority of service members I work with have a disdain (some rather openly) for Hegseth.

“I learned all these skills in the army—smash and grabs, site exploitation—and never got to use them,” he said. “So I’m here to kind of do what I learned to do over there, but this time here, defending my country.”

Never getting to use those skills in real life is the ideal. Any sane person in the military learns a ton of skills that they hope only to use in training. Wanting to use these skills for real, and especially domestically should be grounds for psychologically screening these people out of not just the military but any sort of service position.

Using those skills can be fun in training or drills, actually. So, I get wanting to shoot, run, and do simulated combat. But why would anyone actually want to inflict that harm on someone else for real? They would have to be deeply disturbed.