Posts by QET 67 publicly visible posts • joined 30 Jan 2026
Casually and conveniently ignoring the contributions of, well, contributors.
Which open source projects always needs the blessings of, or 100% certified rewriting instead of ingested & regurgitated by a overglorified algorithm, if attempting something like a license change.
Not to mention prancing around while wearing the skin of its predecessor by not even being a fork.
Yeah, besides, I've always been under the impression that clean room reverse engineering is subjected to what's basically guilty until proven innocent.
And with how all the "AI" crazed companies cranking out models all have scraper bots hammering any kind of publicly accessible servers containing data for "training", good luck proving the the model in question isn't tainted in GPL speak.
Couple it with companies increasing expectations that people they hire (even people fresh out of college/university/whatever) have more experience/skills than what is feasible without working decades in the industry to begin with, and CEO types having found their newest yes-man in the shape of overglorified chatbots ( https://www.theregister.com/2026/03/05/execs_rely_on_ai/ ), it's only a matter of time before reality sets in hard.
It ain't gonna be pretty though.
Each of the Wi-Fi chipsets (2.4 and 5 jiggahertz) in a old ISP provided router I have laying around, has its own OS that gets fed to it upon power-up by the "main" SoC.
Learned that while foolishly trying to unlock the bootloader so OpenWRT could be put on it, instead of it being essentially glorified E-waste.
So much for the middle step of "reduce, re-use, recycle"...
It'd be interesting if that'd essentially make those other AI companies "toxic" on the same level as grok is.
Which could lead to them being considered unfit for public sector or big companies that enjoy a generally positive reputation.
Basically like the whole "cloud" thing and EU sovereignty, except they're just dropped like they're hot instead of replaced.
On one hand, the proverbial monster Elon has been figuratively raising is a perfect fit idealogical wise.
But on the other hand, I have a hard time figuring out how a glorified chatbot who's only good at incoherent right-wing rambling and faux celeb nudies could possibly be weaponized.
If anything, it'd only serve to be a good distraction for the orange manbaby.
But that could ultimately be a net boon...
Those leeching buffoons who were eventually embarrassed had never heard of the concept of a on-prem caching proxy?
I'd have thought any sufficiently large company with IT staff worth their salt would implement something like that.
But on the other hand, companies that big, too often nowadays have beancounters who concluded outsourcing it all could save the company 0.1% compared to the previous on-prem IT staff's salaries.
And that's the core problem of "general purpose "AI"".
It's either gonna suck at everything, need the whole worlds energy output to function (while completely ignoring the actual hardware itself required for it to run, not to mention hundreds if not millenia of years of data scientists carefully crafting the software), or a cocktail of both.
Which once again loops back to the fact that when it's just called LLM and is carefully crafted for a hyperspecific purpose, it's actually great and a boon to society.
But the current hype-train named "AI" which has only a bunch of techbro grifters on-board as engineers, is the exact opposite of that.
Icon for said grifters yelling "ALL ABOARD" to the smoothbrain investors.
The problem here is that Afrikaaner Cartman is both petty & petulant, and stuck in a self-enforcing filter bubble consisting of looney fringe stuff intersecting with hard-right viewpoints.
He was once Tesla's and SpaceX's greatest asset, but is now the biggest liability holding them both back.
In the name of "it just works", never alienate the "prosumers".
Which the mantra and basic design of SystemD does by design with hardcoded dependencies and "obfuscation" (for a lack of better word to describe overcomplicatingly rewriting simple scripts into compiled binaries).
It sure has expedited the process, while also making it much easier.
While what programmers / software-developers actually needed, was a way to prevent themselves from screwing up in the first place, and also a way to easily recover from screw-ups.
But those things require effort and wisdom & knowledge. Things which are the anathema of the "AI" boom.
Lmao... Complaining about Apple users receiving biased news in the OS integrated feeds...
The average Apple fanatic has many flaws, financial irresponsibility being a major one, but being far-right politically ain't one of them, unless the individual is also a tech-bro grifter.
Basically Apple knows what news their core clientelè prefers, and serves that as a result.
As long as you're not the type who likes to stay on the cutting edge of kernel versions, or update to the latest LTS sub-version right away, it's virtually painless.
The snag is that sometimes, things are updated in the kernel which breaks how ZFS "integrates" with the kernel via DKMS. So you'd have to wait for the ZFS specific files for Linux to be updated to compensate for that.
Granted it's been some years since I experienced that myself, but it pays to be cautious with kernel updates with ZFS.
You can't impress intelligent people by dumbing down something so much it's a genuine miracle if it'd even work to begin with.
"fake it 'till you make it" doesn't work in the proverbial trenches of maintaining and keeping the technological underpinnings of the late 20th century and onwards working. It's only good for baffling literal idiots.
And it's the fact that it is represented as capable of doing such repeatedly, despite continous examples of "that ain't ever gonna happen", that's lead to 'em seemingly becoming grumpy old farts, because it is yet another case of "oh boy, here we go again..." of yet another fool who hasn't learned the lessons of the previous fools.
Only lesson here, is that if you can bullshit hard enough to mimic actual scientific progress to the untrained brains of typical "moneybags people" and easily impressed tech clout-chasers, you stand a good chance of separating people from their money.
"AI" is nothing more than the computational equivalent of the whole Juicero farce from around 10 years ago, complete with the equivalent of squeezing the proprietary "mush bags" with your bare hands to render out the juicy contents, is more effective than the overly complex machinery doing it for you.
Especially all the "it's 80% there..." comments from techbros.
Like they got 80% of the winning numbers right for the big lottery jackpot every single time they buy a ticket, while completely ignoring the fact that for there to be 1 winner, there has to be a near unfathomable amount of suckers who lost.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2026
