Note that that site is not official FreeBSD.
I'm aware of that. But I highly valuate
cracauer@'s opinions, thoughts, and ideas.
Frankly I actually don't know about OpenBSD - actually it was more of a thread to leave FreeBSD if done so, and I rate those Open-guys more "fundamentalistic/conservative" in a
good way, so I just
reckon those guys did, or will.
If Linus calls Theo de Raadt "complicated" to me that's a good rating. A sign for somebody staying for a point, and not 'brought round to be open to "some anarchic experiments" anybody can commit to.' (There are points right to do it this way, and points oppose. Or: You can always argue pro or contra. But to do it like anybody else ("swim with the swarm") is always the max BS anyway. (Nobody needs a copy of the swarm when there is the original to join.)
However, if NetBSD did so (thx!), I may switch to it, then.
But I'm convinced so far, FreeBSD will not adapt to it yet.
The pressure of the hype is there, sure. "Everybody is doing it. Why not you? Caveman!"
AI is a technology useful to solve and improve some certain things, if it's used with expertise within its limits - no question.
I use it myself - controlled, and within limits, of course.
But the ([most] heavy) downsides already become apparent.
As you already know yourself, AI can be very nice to relieve you from dull routine jobs - that's exactly what a machine, a computer is all about: You define the picture, but you don't need to do all strenuous handcrafting manual painting work to create it anymore.
Ivan Turkovic wrote some very intelligent, and worth reading articles about that about computers/programming/software development exactly - very recommandable to read. (He's not a AI opponent. I am - within limits, and conditions. He's more professional than me. I may lack words, and maybe I'm also not wrong. Anyway read Ivan {pofessional}! [and me {amateur}])
What I got from my own thoughts, his texts helped me to substantiate (even if he may not intend to. But he thinks clearly, knows software, knows programming, software development, engineering, AI, and its impact on it.)
There is a systematically bug within this AI stuff. (My point, not Ivan's.)
You don't need to be an expert on AI but simply need to know how today's AI by principle works. To know they become stupid if they are not trained (corrected) continously, because they cannot tell right from wrong - they cannot think; they cannot understand.
And they cannot know everything. So in any way, they will do mistakes. Naturally. OK.
BUT:
So, what happens when they do mistakes and cannot tell right from wrong - cannot recognize they made a mistake themselves, and not being corrected by some one who can?
They go stupid.
More and more stupid.
They disintegrate.
That was no problem if there are people knowing better to correct them. But exactly there is the trap.
We saw it with pocket calculators in the 1980s. Math teachers were absolutely right to rebell against to introduce them in elementary classes - but being hushed. Result: The capability of fundamental, most basic calculation dropped to almost zero. (No? So, what's 13 times 7? Without a calculator, in your head, only! Can't do it? My point exactly.)
We saw it at automatic spell checking. Since we write texts with computers with automatic spell checking the amount of wrong written texts exploded.
We germans are known to be overcorrect. That's right. We are trained to be extremely scrupulous bean counting about spelling and grammar within any text of our complicated language. If there was a simple typo in one of the nationwide newspapers it was worth a message in the 8 p.m. news. Until ~ the 1990s. Today you are happy if you can read a single article without any flaw at all.
Point is: People abuse assistance as crutches - just don't walk themselves at all anymore.
Now, back to AI.
Today's older AI users - the grey beards - know programming by writing code by themselves, because they did it - by themselves. So what they use AI for is let it produce the code, then review it with their expertise.
OK.
Now look at somebody learning programming today.
Why learn how to code? The AI does it for you.
You just ask it to write some 3D-shooter written in brainfuck so long until the shit runs as you want it to be - right?
What you wanna review? You don't understand the code. Because you don't learned it. You don't know the language.
Now the computer makes a mistake. And nobody is there to correct it, because the greybeards placing holes into punchcards are died, being laughed at (for placing holes into punch cards), and you cannot handle it, because you never learned it, because you never had to, because the computer did it all for you.
As long as it worked reliably.
So, what once was a tool for the greybeards, who developed it, becomes a crutch, then a substitute, then a replacemant.
Nobody needs to gain knowledge anymore, because the tool povides it.
But the tool needs knowledge to stay reliably.
So, the tool relies on the human's knowledge to stay usably, while the humans trust the tool to provide the knowledge, while at the same time it's going nuts because it's not going to be corrected (trained) by human knowledge.
GAME OVER
I am convinced, this current AI hype will be one of the largest bubble bursts ever.
If there may be no AI-free computer space (OS/programming) left, I toss all that silicon - I focused all my life on since I was 12 - completely into the garbage, and grow plants, bake bread, go fishing, or whatever, as long as if it has nothing to do with computers at all.
Those AIs can talk to each other as long as I don't have engage that BS.
AI is very useful to increase production.
But I don't need to increase production.
I don't want to increase production.
The problem of our society and our planet suffers from, is there already is way too much production.
Why increase it even more?
What do you wanna buy in a desert?
What we lack is quality.
Quality needs understanding.
AI cannot help you on this.
So, where is the point to have AI to increase production, we don't need, while we need to increase quality, where AI cannot help?
There have to be some AI free space left. Where it's not about to produce as much as possible in as short time as possible, but where a intelligent person simply can breathe.
And as I said above, there is a good chance, that a non-AI-infested OSs in middle term may win.
Because I'm not the only human interested in giving up myself for even more production,
but simply want to breath some high quality air, no matter how much more production of air there could be if we unemploy ourselves.
