palordrolap

Yes. The institution in question is human society. We generally grant the permission to make rational decisions over our lives to other humans who know better that we do or are more skilled than we are.

Sometimes, yes, those humans turn out to have been deceitful or dishonest, but there are mechanisms in place for when that happens.

And yes, sometimes those mechanisms are wilfully avoided by the deceitful. Politicians and rich people are especially good at this.

Guess who’s pushing “AI”? The thing that has no contract with human society and cannot be held accountable. And neither will the people pushing it.

This is why we should have as little to do with it - at least as it is in its current form - as possible.

This is true. But then I’m not using the latest version while I still have an active session, and that can lead to weird behaviour or errors after the fact.

Case in point, I once received an Xorg update that I allowed to install, but didn’t restart the computer properly until much, much later.

By then I’d forgotten about the update, so when I restarted and started having graphics problems, I was mystified.

I’ve also forgotten how that all panned out, but in the same situation I’d roll back to a previous Timeshift snapshot and work the system forward again until I find the culprit or things are stable, so I assume that’s what I did back then.

For me, it’s about reducing the amount of time the “update available” icon shows up in the system tray, because its very presence bothers me. Maybe there’s something cool and new. Maybe it fixes a severe security problem. If it’s for programs I’m not using right now, then the update can be applied right now. Otherwise it’s going to have to wait until I’m done. And bother me.

Yes, I could turn updates off and never see it, but that seems like a bad plan in the long run.

Fake AGI is like fake banknotes. Some of them are really good approximations. Nigh indistinguishable. A lot of people will be fooled by it but eventually it will be discovered to be a fake and people will get hurt in some way or another.

And it won’t be the people who are pushing for “AGI”.

The people who are seeking AGI will be happy when an LLM appears clever enough to fool them, not anyone else.

They may even realise this, because they think everyone else is less clever than they are.

This is why the whole thing has been called AI in the first place.

As I’ve said before, once Linus is gone, we might well end up with splits at the kernel level rather than at the distro level. And we would be wise to avoid any one organisation’s stock kernel, even if there are some very large organisations providing a lot of code for the kernel at present.

I can see a future where, say, GNOME, start producing their own kernels to support their vision of the Linux desktop from the ground up.

And it’s all but certain that Canonical and Red Hat would be very interested in things going their (respective) way(s) when the time comes.

An old computer trick / prank / “fun” thing to do was piping random things to /dev/audio, or finding whatever program was available that could take any old file and not complain while translating it to audio by some means or another.

On my distro there are at least three of these programs installed by default: aplay, paplay and pw-play.

Some or all of these will complain if the file or stream they’re given isn’t a recognisable audio file, in which case, there’s a --raw or similar flag where it’ll just shrug and blast whatever through the sound system. If you’re creative, you can set different sample rates and hear it at different speeds.

VLC is just a really fancy way of doing the same thing.

For even more “fun”, try opening a file in Audacity / Tenacity, which will default to raw mode if it can’t tell what a file is, and you get to see the waveform and so on. Just take care not to modify and save over an important file with that.

Any C-levels, current or former, who were in charge during the unethical behaviour should be made to pay this. Either they knew or they should have known. Seize their assets, take them back to basic income and lifestyle and only then take the remainder of the fine from the company.

Yeah, I know this is a pipe dream. These people have their arms lodged so deep in politicians it makes their mouths flap, so it’ll never happen, but it would stop this sort of behaviour right quick.

Yeah, my university had those, but they also had an interface to it accessible from the more modern systems.

I also did a work experience placement with a company that had amber-screen terminals when I was still at school (and the year still started with a 1), so I’m no spring chicken either. They were very early in the process of supplanting them with PCs, which is not something they explicitly told me, but looking back, the evidence was all there.

The “fun” part with those specific terminals was that the admin password for the terminal hardware itself - because they had a rudimentary sort of BIOS on them - was a “fail at the first wrong character” system. With enough tries you could figure it out.

There wasn’t much you could do from there, at least not that I remember, but one of the terminals I used did end up beeping at a slightly different frequency to all the others.

A terminal in the computer sense was originally a screen and keyboard attached to a terminating node on a network. The network didn’t pass through, so it terminated there. This meant the literal, physical hardware. Think old school green- or amber-screen systems attached to a mainframe in the basement somewhere.

A console was a terminal that was serving some kind of purpose and showing some kind of interface for humans to interact with. Without the interface software, a terminal is not a console. Without the hardware, you wouldn’t have either.

It’s easy to see how these things became blurred.

And now it’s worse because we’ve extended the meanings a bit. The program in our fancy GUIs called “Terminal” and which we often just call “a terminal” is actually a terminal emulator.

And to a lesser extent, so is the thing you can access on many distros by pressing Ctrl+Alt+F1. This sometimes gets called “the console” because it’s even more like those old terminal interfaces. Full screen. Text only. Largely monochrome. No GUI.

And deeper still, a terminal, console, or terminal emulator doesn’t have to mean “a shell” which is another thing entirely. Shells just happen to be one kind of interface that can run there, and is often the default option in a GUI terminal emulator.

From a console, the default program is generally some flavour of login prompt. And then the system automagically loads whatever is configured as that user’s shell once they log in.

I was going to suggest using pro-ICE rhetoric dripping with over the top sarcasm, but there are too many sufficiently smart pro-ICE people who’d see through it and too many anti-ICE who might have trouble discerning it. And then there’s the issue of getting the tone right in a stressful and dangerous situation.

… so you don’t think Mastodon is a worthy competitor?

How about Bluesky?

I made sl on my computer a bit more literal. It takes the output of ls -l and reverses every line, including any wrapping within the column width, and pads it to the right of the terminal. One day I might get around to fixing it so that it forces, parses and correctly reverses the ANSI colour codes too.

In /usr/bin, I get lots of lines that “start” with spaces and “end” with things like toor toor 1 x-rx-rxwr-

US English dialects mainly, though there may be pockets in other Anglophone places.

Long was it known fact: Windows versions and OG Star Trek films. Every other one was terrible.

… but I note there are a few important releases missing there. 3.0, Win2K and 8.1 especially, and we might argue for 3.1 and 98SE and maybe even the unreleased Longhorn too.

Using AI to find errors that can then be independently verified sounds reasonable.

The danger would be in assuming that it will find all errors, or that an AI once-over would be “good enough”. This is what most rich AI proponents are most interested in, after all; a full AI process with as few costly humans as possible.

The lesser dangers would be 1) the potential for the human using the tool to lose or weaken their own ability to find bugs without external help and 2) the AI finding something that isn’t a bug, and the human “fixing” it without a full understanding that it wasn’t wrong in the first place.

The basic functionality of sponge can be emulated with an AWK or Perl script, so most people who needed it in the past almost certainly rolled their own.

I get what they’re going for with the arrow coming from the process to STDIN, but I still feel like it should point the other way.

And shout-out to the sponge and tee command-line tools for those situations where the memory buffer won’t cut it.

Is this the first human trial, or just the first officially sanctioned one?

IIRC there was that one guy who experimented on himself and cured his lactose intolerance.

… found it: https://www.youtube.com/watch?v=J3FcbFqSoQY

Nearly 8 years ago.

“I like to rebuild my kit sports car every time I want to take it out for a drive. Anyone who does otherwise is a pleb.”