Don't connect critical infrastructure to the internet.

And yes, food safety is critical infrastructure.

But then, when do you actually have to be at work? At 8 UTC? Which is somewhere after dark where you live?

In practice, audit trails are 99% about reporting and monitoring needs, not liability after-the-fact. By the time the insurance companies are involved, you're already tossing thousands of dollars of food and possibly have sick people lining up to sue

If there was any sort of centralized monitoring involved here, they would have caught the problem as soon as the temperatures rose above $AlarmHigh or remained above $AlarmLow for more than X minutes. They did not. The exploited interface was therefore a function of convenience, not compliance.

Then you should have no problem identifying infringing code?

AI is technically accurate. ML is literally the core component of the 1955 conjecture that defined AI. LLMs are a display of intelligence as a result of ML, also directly part of the original conjecture.

Sorry, no.

LLMs do one thing that was considered probably to be one of many attributes of what the proposal authors conjectured might be required to qualify for their definition of "artificial intelligence". But they do none of the other things: they do not form abstractions or concepts, they do not self-improve, they do not display human-like creativity. They only generate a statistically-probable stream of output tokens rendered as human vocabulary. If that counts as "a display of intelligence" to you, then I worry about what you think intelligence requires. Was ELIZA "a display of intelligence"?

The original proposal also did not define the modern understanding of "machine learning" as part of its outline of "artificial intelligence". The closest it got was this: "...admitting a new word and some rules whereby sentences containing it imply and are implied by others." But in the very next sentence the authors note "This idea has never been very precisely formulated nor have examples been worked out." What is commonly understood as ML by practitioners encompasses a universe of much-better-defined techniques than the first sentence quoted above, including some that don't fit that sentence at all. What was done with Cyc, feeding it carefully-crafted rules of language and logic, would fit into the above definition, but is not ML, and a lot of ML activities like visual object recognition wouldn't fit the above definition because they have nothing to do with language per se.

More or less, yeah...

You can use ArcaOS, you can use CSM mode, or you can chain-load it with any of a number of EFI loaders.

CSM mode is probably the closest to "pure" OS/2, but obviously it's not really using UEFI. ArcaOS is the closest to pure UEFI support (and probably the best choice if you plan to really use it rather than just prove you can do it then move on), but it's not pure OS/2. Take your pick.

/ I am not affiliated in any way with ArcaOS / Arca Noae.

It's not correct for Germany, but it is for Austria. Austria was driving on the left side of the road until the Anschluss, but switched over to right side drive to harmonize with Germany.

I wounder if AmoredDragon can come up with a similar slur to Douchelanders for Austria.

Foreign students need a student's visa, or to be more specific, an F-1 visa for colleges and university. And before they get that, they have to bring equivalent documentation to the U.S. embassy in their home country or country of residence.

I hate Zuck as much as the next geek, but preserving human rights is a governmental-scale responsibility, not an issue we leave up to every random company to do their own due diligence (or we'd have none).

It's great Wyden and Markey are taking an interest in this, but the issue is far bigger than one major brand of spy-glasses; there are dozens of brands of similar glasses that don't even pretend to try to protect privacy; and moving beyond that specific form factor, there are thousands of styles of spy cameras available to any random Joe for under $100 (many in the $20-$30 range!) right from Amazon. This isn't even trying to squeeze the toothpaste back into the tube - This is drowning in a vat of Colgate while pissing and moaning over a stray blob of Crest.

Two engineers (one of whom is a stuffed shirt being thrown a bone post-acquisition) are going to rewrite Cisco's legacy code using AI and an unlimited budget.

I'm sorry, can even the most delusional pro-AI executive out there seriously say that with a straight face? This just plain fails the sniff test - Are they establishing a scapegoat for an already-failing project? Honoring some contractual landmine EZDubs managed to sneak in? Straight-up insurance fraud?

If this is exactly what it looks like at face value, excuse me while I go short Cisco stock.

> I'm not saying LLMs will not write security holes, they definitely will if you don't very explicitly tell them not to. If telling an LLM "Make no mistakes" actually worked, you wouldn't need to tell it that in the first place.

> > 2) contribute slop patches to existing open source > And open source projects are increasingly disallowing AI generated code due to copyright problems. And those projects are finding that a large number of people still submit LLM slop patches anyway. Even if they're all rejected immediately upon submission, it's still a burden on the project. As a result an increasing number of projects are limiting the ability to contribute, some basically closing outside contributions entirely.

Moderators are a special kind of special.

I got banned from the entire site for "circumventing" subreddit bans, which was because some of my alts were fed subreddits by the algorithm that I was not supposed to be on.

...Nothing on them but cranky bastards contributing to climate change while moving between ticky-tacky, 40+ hours of indentured servitude, and temples to Mammon.

I largely agree with most of Klee's points, but there's a serious bias issue with his conclusion: "The internet" is just the road we use to get somewhere. Where we go is up to us most of the time. And in this case, the "old internet" is still mostly there - Case in point, I still have a crappy static personal website up that hasn't been updated since the 1990s.

Problem is, nobody wants to read my ramblings from 30 years ago any more than they did 30 years ago. People want FAANGs, even while they may openly resent those same platforms. Sure, Amazon is evil... But the alternative is leaving the house and going to a store, eew! NetFlix kills everything good after two seasons, but two seasons on-demand beats a 30-second stick-figure animation that takes six weeks per "episode" and depends on the unpaid hobbyist creator not getting bored with the project. TikTok is measurably destroying the attention spans of our youth, but we can't risk missing the next big "hold my beer and watch this" moment.

The problem isn't "the internet". The problem is us as a collective whole. When there's demand for trash, sanitation engineers flip to supply side.