I can see the challenge in sorting out AI slop from actual art or writing being normalized in the same way that occasionally having to check your spam filter in case an important work email got filed alongside "GrOwYoUrEgGpLaNtEmOjIfOrChEaP", but there's a difference between a world where AI slop exists and AI slop itself actually being worth a damn.

I feel like you're missing an important piece of context. The "nobody reads the article" joke works (when it works) because it's both a known common behavior and one that creates incredibly obvious problems and makes people look silly. Apologies that it was your turn to take that L last week, I guess?

A fair chunk of this also comes down to community norms. On awful.systems it's not really an expectation that the top level post includes a summary instead of just some choice quotes or a reaction, because the expectation is that the thing being shared is worth reading in full so of course you're going to do that if you have time. Hell a lot of our members are sometimes sharing either their own longer-form work or that of people they know personally and it'd be rude to act like following the link and reading the thing isn't worth doing. I'm not arrogant enough to say I always read the link before commenting, but unless I'm making a bad pun off of another comment or otherwise shitposting I usually do, and it's kind of baffling to have someone actively defend doing otherwise.

And yes, when someone uses the phrase "my dude" you can assume they're at least 65% taking the piss.

Musk is the most boring and pathetic kind of unfunny where he desperately wants to be in on the joke but is terrified that the joke is on him (because it is). Rather than accept this with any kind of humility he instead cannot accept the L and has basically spent all his vast money and power making that everyone else's problem.

He is the worst mad scientist, ranting about how they called him mad when what we actually said was "lol u mad bro?"

Wait, do you mean new Armenia or Old Armenia? I'm not fluent in the region. But I'm sure Turkiye will be happy to surrender territory to Old Armenia after denying that whole affair for the last ~100 years.

Hang on I've got to reread this whole thing in a skeletor voice now.

I mean, I suspect that when he says he wants someone else's genes he probably means Elon's, so they probably wouldn't actually disagree over much. Because he's successful and success is determined by genes and therefore he has the best ones, you see.

Sweden/Svealand is one of many examples here of "where do you think the name came from?" for neighboring regions. My favorite example of this pattern is probably Finland, though. Helsinki is it's own thing, and we get both Finland and Finnmark. Truly outstanding.

I like how the Spanish coast is in Africa. Also, Kyiv and Helsinki are their own city-states (or autonomous post-state administrative zones or whatever) but Rome, London, Munich, etc. aren't.

Im sure they'll try to incorporate an LLM into the stack somewhere, leading to at least one car that's exposed to the "pretend to be a fire truck" attack.

I think the idea is to solve that by networking all the self-driving cars together. I'm sure the long history of trying to get vendors to agree on a standard when they all benefit individually from the lock-in of proprietary systems has nothing to teach us about this prospect.

I don't know. Some walks are very distinctive.

I mean it fits into the pattern of procurement projects that aren't allowed to fail despite having had serious coherence issues starting at the design stage. Though the military is usually less prone to the "problem in search of a solution" dynamic that VCs are prone to if a project gets started it can shamble forwards as a zombie for years before anyone finds the political will to kill it.

Hat tip to the commenter over there who says point-blank "this is just eugenics."

Also gotta love the Rat community treating very contentious assumptions like settled fact, particularly re: heritability.

You call it a problem. I call it a O(1) mining algorithm.

Extreme poverty in particular is a far less objective or meaningful measurement than you would think given how often the "everything is fine" crowd likes to cite it. The daily income defined as "extreme poverty" is abysmally low; $2 USD per day wouldn't be enough to get basic necessities for food and shelter, and while its terrifying to think about having to live on even less we shouldn't congratulate ourselves when the bar is just barely above the lowest levels of hell. Different poverty lines show different trends and by standards that would allow a person to live decently rather than merely avoiding the absolute worst deprivations we actually see very little change. And that's before getting into the way poverty is distributed globally and the ways that even with the lowest poverty lines we see a lot of the poorest of the poor who have seen far less of a shift.

I mean, according to the charges Tate hiring young women usually meant some variety of sex trafficking and adult video that he took the money for. Tbh the whole space is sufficiently toxic that she ought to start dropping the banhammer judiciously, but IDK what situation is politically economically etc.

I'd bet dollars to donuts that the internal documents from OpenAI on this marketing push are pretty clear about the real goal here. Plagiarism is one of the most visible and easiest to understand problems enabled by GenAI. My wife is getting an online degree and it's incredibly obvious how many other students are just shamelessly dumping the assignment into chatGPT. So they need to reframe it as part of a wider conversation about GenAI and education, which is where you get the nonsense buzzword courses that don't attempt to engage with even the most obvious problems.

I like how he even had someone with art expertise literally explain it to him and he writes it off as "lol she must have super artist vision for details."

I don't know there's something here about how broken the way we engage with art is. How commodified art is inherently decontextualized and while you can see the beauty or the power or whatever you lose something without the curation and presentation you get from a gallery or a museum.

I also want to dunk on a few of the specific inclusions. AI clearly doesn't understand the point of cubism in particular, making it an exceptionally clear example of what Scott's artist friend was talking about. Including a digital photograph of a collage that clearly makes use of the depth of the actual work is pretty dumb.

I heard the Chinese have a grooming machine that... actually I don't think I want to finish that joke.

See, isn't the 4-hour work week one of those "just make other people work 50+hours a week on your behalf and take the money they've earned for it" schemes? This looks much broader rather than being married to a specific sub-scam. Like, if crypto is down they can sell drop shipping. If drop shipping is cringe they can sell AI slop monetization. If Amazon tightens their standards and starts locking out AI stuff they can go back to crypto.

It's in the same genre of trying to monetize being a conspicuous asshole, but it is one of the more complex evolutions, at least compared to the standard grift-luencer.