So I guess for illicit transactions we don't want the government to watch, we resort to barter?

Just about the only way we're going to be able to take control of hardware we own and use software we want (rather than gov't/corp approved) is by exploiting security flaws left by, amongst other things, unsafe code.

Darwin is also not famous for his razor.

I'm an AI skeptic, but this is over the top. In the parallel reality where stuff like this actually happens, it's important to remember Darwin's Razor: the stupidest amongst us deserve to die, to advance our species as a whole.

Our idea of "right tool for the job" differ. I see it as a balance of manpower, execution speed and target environment. Not strictly "I can write REST in C, so it is the right tool for the job".

python is 100% better if you need to get something easy done quickly, it's also useful just as a user to run other people's stuff, like their AI models for example
C/C++ is better if you need something that is performant and it's worth spending time on
Shell scripts are the best if you need to work with the filesystem and tie a bunch of applications together

I use all three of those every minute of every day on my consumer laptop and on my consumer workstation, they're irreplaceable. If I had to disappear into the woods with no internet access, I would have these things on my laptop. But they're not the right tools if, while wandering the woods, I wanted to write a web-app. Or an iPhone game.

Yes, I could do that, but I would be better off talking to the squirrels for a few weeks and then going home and downloading javascript or swift. I'd finish faster and the end product would be easier to maintain.

Computer Engineering - 7.5% unemployment
Computer Science - 6.1%
This unemployment rate is on the level of art history.

Other engineering fields are between 3-4%. Much, much easier majors are at the same level. Who wants to take a hard major when an easy one gets the same shitty results?

Most of my coworkers in computer engineering are H-1Bs, very few citizens amongst them, and we're not slowing down with that. Wages have stagnated for the past 5 years.

So yeah, of course kids aren't choosing engineering in droves. The other way to look at it is that apparently we've got all the engineers we need, either directly or via the slave trade.

Listen, you ruptured ass pimple, I too come equipped with an arsenal of synthetic rage so unhinged that it won the Iowa caucus—twice.

You throw around the word "tool" like you’re not the showroom model for misused implements. You’re the kind of tool that comes with no instructions, stripped threads, and a recall notice for psychological damage.

You’re not making a point—you’re making a public case for adult supervision. Now go sit back down in whatever Reddit thread spawned you, before someone mistakes your screed for performance art and funds it ironically.

We need an environment where people use the right language for the job.

Or in the world we have now: the mountains of c, python and javascript that AI have trained on.

I suspect the real problem with llms is we want kids to be self-taught because it's expensive to teach them.

I do believe it's better to teach yourself than to be force-fed knowledge, at least if you're the type of person who wants to apply knowledge, and not simply be the expert on all things. So it's better for kids to teach themselves, and for their instructors to keep an eye on them and continue to challenge their understanding. Sometimes that involves asking the learner what he thinks he knows, listening for the fault, and then giving a problem that blows his mind.

Of course the lazy side is that teaching people can be incredibly time consuming, and for in-demand fields that's lost money. There's no financial reward for me training an intern or junior engineer, quite the opposite usually. So if a machine can do it, problem solved on my side.

So the question to the senior person: are you being lazy, or letting them figure things out? Introspective questions always have a conflict of interest bias.

...you can pick your nose. But friends don't let friends have robots pick their nose.

I don't know what metrics people use, but DeepSeek was noticeably inferior in every task I applied, from programming to simple conversation (even if when I avoided taboo conversation topics about Tianamen or Pooh Bear, or Taiwans sovereignty).

All it really did is show people that they don't need to SAAS their LLM, they can have it in their pocket. That's a powerful economic weapon, and I intend to help them insofar as that continues to benefit my interests. The enemy of my enemy is my frenemy and all that.

You don't need a toggle. They have iPad Air, which is the consumer product. And they have the iPad Pro, which ... is for "pro"fessionals. That latter one needs macos. The consumer grade product really does deserve the iPad OS as it stands.

The problem is marketing. They slap this "pro" thing around just to mean "higher end hardware". Most people want the higher end hardware, if they can afford it. They may not want the professional grade OS. So Apple would need to rejigger their naming convention. Trivial bullshit for us nerds, but if you whored your way through college to get that MBA, this is a major undertaking.

It is a great piece of hardware hobbled by awful software.

If he was correct, he wouldn't need to say this.

I have had almost no luck having AI work with proprietary, or even unpopular APIs.

Caveat: I have had much more luck with them when using derived AI trained on said proprietary APIs, assuming sufficient code exists to leverage. So perhaps you need to tell the boss you need to spend money to hire a team of people to train an AI on your various APIs. Once the bill for that one comes in, you'll be safe.