FaceDeer

Things change. There was a period before this information was easily available; this repository only goes back to 2013. Now there’s a period after this information, too. Things start and eventually they end.

Here’s hoping that some neat new things start up in its place.

They’re not both true, though. It’s actually perfectly fine for a new dataset to contain AI generated content. Especially when it’s mixed in with non-AI-generated content. It can even be better in some circumstances, that’s what “synthetic data” is all about.

The various experiments demonstrating model collapse have to go out of their way to make it happen, by deliberately recycling model outputs over and over without using any of the methods that real-world AI trainers use to ensure that it doesn’t happen. As I said, real-world AI trainers are actually quite knowledgeable about this stuff, model collapse isn’t some surprising new development that they’re helpless in the face of. It’s just another factor to include in the criteria for curating training data sets. It’s already a “solved” problem.

The reason these articles keep coming around is that there are a lot of people that don’t want it to be a solved problem, and love clicking on headlines that say it isn’t. I guess if it makes them feel better they can go ahead and keep doing that, but supposedly this is a technology community and I would expect there to be some interest in the underlying truth of the matter.

No, researchers in the field knew about this potential problem ages ago. It’s easy enough to work around and prevent.

People who are just on the lookout for the latest “aha, AI bad!” Headline, on the other hand, discover this every couple of months.

AI already long ago stopped being trained on any old random stuff that came along off the web. Training data is carefully curated and processed these days. Much of it is synthetic, in fact.

These breathless articles about model collapse dooming AI are like discovering that the sun sets at night and declaring solar power to be doomed. The people working on this stuff know about it already and long ago worked around it.

This is “technology news and articles?”

Seems like this place is increasingly just people yelling at AI-generated clouds.

Sometimes headshots develop spontaneously. It’s a rare condition, but convenient. Some claim John F. Kennedy suffered from this condition.

I recall seeing a list of the most dangerous jobs in America and “President of the United States” topped it due to the high percentage of people with that job who’ve been shot.

But at least that crappy bug-riddled code has soul!

In Tyreek’s post-arrest press conference he asked rhetorically “what would have happened if I hadn’t been famous?”

Well, now we see. Wrist-slaps with no actual long-term impact.

It’s almost doublethink, people celebrating how the Fediverse is an open protocol for sharing public discussion and then going surprised-Pikachu at the notion that public discussion might be viewed by someone the don’t want to view it.

If you don’t mean for something to be public, don’t post it on a public forum.

I’m Canadian so I’m not a voter in the contest you’re presenting, but if I were I would vote Democrat. And of the trio you present for the Democrats, I would say that the position I’d compromise on would be gun control. Not because American gun culture isn’t bananas and it’s not a serious problem, but because I can’t see any plausible way to fix it in the short term. So might as well let it go for now and deal with the more important stuff that affects more people.

I think a more reasonable compromise would be to give Republicans most of what they want on immigration reform. That seems to be something they consider to be of critical importance, but that I think can be allowed without it causing significant harm. If the American economy starts to suffer as a result of not having illegal immigrant workers then that will be motivation for further reforms. I think it’s important to have the laws try to reflect the realities, though, and having the economy literally depend on large-scale lack of adherence to the law of the land is a bad place to be. Just make sure not to be monstrous about it - don’t do the concentration-camps-for-children thing, try to maintain basic asylum access for those who truly need it, and so forth.

Yup. I would personally love it if the electorate studied the various policies of candidates that wished to be their representatives, decided which ones’ positions were the most beneficial to themselves and to the country as a whole (which is indirectly beneficial to themselves, after all) and then selected that one on a rational basis. If we lived in that world then each candidates’ campaign would ideally focus on debating issues and presenting their views.

We don’t live in that world, alas. I’ve become cynical about democracy of late because the electorate are a bunch of sports team fans who just want “their guy” to win. Well, so be it then. It’s kind of an emergency right now so play whatever strategy keeps the regressive loons out of power.

It’s often not a choice between an AI-generated summary and a human-generated one, though. It’s a choice between an AI-generated summary and no summary.

Could it perhaps be that online communities are in bubbles that focus primarily on his failures and downvote into oblivion any mention of successes he might have had?

…

No, it must be the money that’s wrong.

Okay, now hit it again.

Indeed. And Boeing is the main contractor for it so you can be sure it won’t suffer any mishaps.

There’s others that are trying, Blue Origin has their New Shepherd rocket that is able to land, but it’s a suborbital tourism vehicle that’s basically just a toy. They’re working on a partly-reusable orbital launcher that’s like a souped up Falcon 9 but it’s still in development. Several other smaller startups are working on smaller Falcon-9-like launchers with expendable second stages, and China is building a straight up carbon-copy of the Falcon 9 and Starship. But SpaceX is the leader in this field and currently the only one who’s actually successful. Everyone is following in their wake at the moment.

Indeed, I’m surprised this dumb clickbait title didn’t literally include Elon Musk’s name like so many other “Elon Musk’s <Company Name> Does <Thing That’s Actually Normal But Sounds Bad>!” headlines.

Yes, Elon Musk has some awful views and does some awful things. Doesn’t mean everything he does is therefore bad. Henry Ford was a colossal antisemite, as another example, and did some really weird and awful things to his employees. Unfortunately some of the same personal characteristics that can lead people to be innovative industrialists can often also lead to them being assholes.

So now it’s basically people who aren’t going to use this tool complaining that other people who do want to use this tool will get to use it.

The chart is actually not so terrifying once the cause of it was discovered. Basically, new regulations reducing sulfur content in ship fuel that went into effect in 2020 and put an end to an unintentional geoengineering experiment that had been running due to the enhanced cloud cover the sulfur dioxide particles were causing.

So this is a strong data point showing that solar geoengineering could work, if we resume it deliberately under more controlled conditions.