I’m not trying to be overly self-promotional, but rss.ponder.cat + piefed.social’s “feeds” could probably do this very well. I actually really like the combination of having aggregated RSS feeds I can control, along with other people voting on them so particularly interesting stuff from them get surfaced more than other stuff.

I was careful with my phrasing for exactly that reason. I think that, either, before the end of his current term things will blow up in such cataclysmic fashion that all bets are off for what comes next, or else Trump will be starting his third term in 2029.

As much as I am scared for it, I am hoping it is the first.

Not even Elon Musk gets to ignore Wikimedia policies. That will never change.

I hope this is true. By the end of Trump’s term, ICE may be able to simply storm the Wikimedia offices and shut everything down if they don’t adhere to Trump’s official truth.

I assume they have enough international presence that things will be able to continue, but their core office being in the US is no longer the pretty-good protection against authoritarian regimes trying to force them to alter what they’re presenting. I’m sure they will fight, but a lot of the tools they used to use to fight with are being attacked, with quite a bit of success, right now.

Wikipedia is extremely resistant to this kind of thing. They’ve been dealing with it for decades now. All kinds of people have had that idea, sometimes professionally being paid by some company/country/individual to the tune of quite a bit of money. Sometimes they do succeed in bending the narrative slightly in particular ways, but the simple “overwhelm everyone’s resources” approach that works a treat on Lemmy or Facebook can get simply brushed aside on Wikipedia because of its strong culture and good design features.

Ah, I got it. Yeah, it makes sense, WP.com is moderately likely to keep working fine probably, it’s just that it would make me nervous at this stage. I just don’t think he can do anything to really “punish” Bluehost if they’re using his software in some way that displeases him. WPEngine’s mistake was getting tangled up into a business relationship where they were depending on listings and APIs and things. Although, it probably seemed like a good idea until their business counterpart went off the deep end.

It’s a pretty hard thing to hold back the tsunami of assholes who descend on a company once it starts making above a certain amount of money, who want to use it for evil.

Yeah but why would the company run by the crazy person be the only safe place?

It’s open source. Just find a different host that isn’t run by a known unstable human. Literally any other. That would be my feeling on it, at least.

Yes yes this is a very good point, stay well clear of Wordpress.com, Automattic, or any similar nonsense. All I meant by “Wordpress hosting” was managed hosting from some third-party place like Bluehost or Hostinger. The software is fine, it’s all open source and the worst that will happen is 6 months from now, it’s not getting a lot of feature updates because the core company that was making it has imploded completely, and someone from the community has taken over security updates.

But yes you need to stay clear of the clusterfuck while it’s going on. Don’t use Wordpress.com or anything adjacent to it.

Edit: Wait, I didn’t even read closely enough. Why would Wordpress.com be safe? I had some vague impression it was connected with Automattic in some way, although I’m not sure, maybe it is just one of the third-party companies. I just feel like anything that’s in any way adjacent to Automattic or anything “official” about Wordpress would be best avoided for a while.

Yeah. I’ve run plenty of services from a computer sitting in someone’s office, or in my living room, while they’re in-production-while-in-development. Sometimes it makes sense. But it’s just not something you want to deliberately aim for as the solution. What if the power goes out? What if your motherboard dies? What if the toilet overflows when you’re not there, and floods the place?

Just get a dedicated service and pay them their $10/month and have them worry about all that crap for you.

It’ll be vastly cheaper and easier to just get hosting somewhere.

Wordpress hosting (edit: THIRD PARTY Wordpress hosting, Bluehost and Hostinger are decent I think, see below) is fine for most small businesses and starts at about $10/mo. You can go fancier and more reliable and go up to $30/mo or something, or if you really need your own VPS you can go with Vultr or Hostinger and get a pretty similar price range for pretty much whatever you want to do.

I think the only reason to self-host is if you have some crazy special hardware or legal issue, or your own dev stuff that you don’t want/need to push to “the cloud” to put it online. Otherwise it’s such a buyer’s-choice market that it’s hard to justify.

Hey, can you summarize the core argument of this for me? Specifically with an eye to its broader implications about the AI sector and the technology industry as it currently stands in its current form today.

I’ll analyze the core argument of this piece and its broader implications.

The article’s central argument is that DeepSeek’s recent breakthrough has exposed fundamental flaws in how Western tech companies, particularly OpenAI and Anthropic, have approached AI development. The author contends that these companies weren’t actually innovating but rather perpetuating what they call a “Rot Economy” - a system where companies burn through massive amounts of capital while creating an illusion of inevitable progress and necessary scale.

The key revelation isn’t just that DeepSeek created more efficient models, but that Western AI companies apparently never seriously tried to do so. The author argues this wasn’t because they couldn’t, but because they had no incentive to - they were operating in a system that rewarded spending and scale rather than efficiency and innovation. They had convinced investors and the market that the only path forward was through ever-larger models requiring ever-more expensive infrastructure.

The broader implications about the current state of the technology industry are particularly damning:

1. The modern tech industry, particularly in AI, has become fundamentally anti-innovative. Rather than solving real problems efficiently, it’s oriented around maintaining narratives that justify continued capital accumulation.
2. Silicon Valley has lost its “scrappy builder” spirit. Instead of finding clever solutions to constraints (as DeepSeek did), major companies simply throw money and resources at problems.
3. The venture capital and tech media ecosystem has become complicit in this dynamic, uncritically accepting narratives about necessary scale and inevitable progress while failing to ask basic questions about efficiency and practical value.

The author suggests this represents a deeper crisis in American tech innovation - the industry has become dominated by what they call “management consultants that lead teams of engineers to do things based on vibes” rather than genuine technological problem-solving. This has created an environment where companies compete not on actual innovation or efficiency, but on their ability to raise and spend capital while maintaining compelling narratives about future potential.

The piece concludes that much of the current AI boom has been essentially a confidence game - not because the technology doesn’t work, but because the companies developing it have been more focused on maintaining narratives that justify their continued fundraising than on creating genuinely valuable and sustainable products. DeepSeek’s achievement matters not just because it created more efficient models, but because it demonstrated that the emperor had no clothes - that the massive infrastructure and spending of Western AI companies wasn’t actually necessary for technological progress.

This suggests a potential inflection point for the tech industry, where the narrative-driven, capital-intensive model of innovation may be reaching its limits, potentially forcing a return to more genuine technological problem-solving and efficiency-driven development.

My AI summarizer is superior to your AI summarizer. 😃

He just does whatever the last person to talk with him was able to talk him into.

I suspect that, at the root, he just likes being popular with people and he has some kind of understanding that everyone who doesn’t look any harder than “But I like going on TikTok, what the fuck” will get mad at him, and so he’ll undo the ban. But it could honestly go any which way, just depending on the whims of whatever little hamster is running on the wheel in his head on any particular day.

Yeah. Techdirt has been saying some weird stuff lately. They threw a fit because Biden said he wouldn’t try to enforce the TikTok ban over the weekend, but was instead handing the ban they fought tooth and nail for to Trump, on Monday, for him to enforce. As pretty much any sane person would assume to be the case. They said things like, “This whiplash-inducing reversal from the Biden administration, after championing the TikTok ban, underscores the arbitrary and politically motivated nature of this decision.”

McVoy first blustered and threatened, but ultimately chose to go home and take his ball with him: he withdrew permission for gratis use by free software projects, and Linux developers will move to other software.

If I remember it right, he did a lot more than that. He tried to say that one particular kernel developer who he viewed as disobedient to him would be punished by no longer being allowed to use the software. When people pointed out that this behavior was insane and would cause significant disruption to the project, he didn’t care. Then, they made the absolutely predictable choice to abandon him. Then he took his ball and went home, after everyone had already moved to a nearby park and started a new game without him.

I might be misremembering, but that’s how I remember it happening. Instead of using git, we could all be using BitKeeper, and paying McVoy our $5/month or whatever for the privilege, because it was just as much better than everything else as git is now. But he didn’t want that, if it involved not having everything exactly the way he wanted it.

I know of no faster way to relegate your project to the dustbin of history.

It happened with X. XFree86 was the graphics system you used on Linux. One developer had constant friction with the core XFree86 people, but he was also a guy who kept coming up with good and innovative ideas and making them happen, and had a lot of respect from the wider community, and so for a long time there was this uneasy tension. Finally, things came to a head:

https://www.zdnet.com/article/dispute-divides-key-open-source-group/

I think it took about a week after that before Keith was leading a new core group of developers and sensible people, and everyone was simply totally ignoring XFree86. All the distributions switched to Keith’s fork, xorg, which they continued to use for about 15 years, until Wayland came along.

It stands alongside Larry McVoy telling the Linux developers they needed to jump through hoops to use his version control system, because they had no alternative, in the absolute hall of fame of completely unforced own-goals that changed the landscape of software in ways that are still felt today.

Edit: Typo

Russia: We’re so mighty that all must tremble before us, our economy is stronger from sanctions, NATO is on the verge of collapse

Also Russia: Ma make em stop they’re being mean to me on Mastodon

Real gangsters don’t sweat the jibber-jabber.

It’s hard for me to tell if I’m just set in my ways according to the way I used to do it, but I feel exactly the same.

I think Docker started as “we’re doing things at massive scale, and we need to have a way to spin up new installations automatically and reliably.” That was good.

It’s now become “if I automate the installation of my software, it doesn’t matter that the whole thing is a teetering mess of dependencies and scripted hacks, because it’ll all be hidden inside the container, and also people with no real understanding can just push the button and deploy it.”

I forced myself to learn how to use Docker for installing a few things, found it incredibly hard to do anything of consequence to the software inside the container, and for my use case it added extra complexity for no reason, and I mostly abandoned it.

It’s part of a longstanding tradition of abandoning our less powerful allies once their usefulness to us is at an end. Our South Vietnamese friends, then the Afghans, then the Kurds, the Iraqis, the Afghans again, and now I’m sure there are some people in Syria whose day is coming due.

The difference is that Trump is planning to do it to everyone, on purpose, before their usefulness is even at an end, for no reason at all. Not just people who were forced into their alliance in a desperate time of need, but people who were doing perfectly fine in their civil society in whatever corner of the world, who opted on their own to help us out, are now going to be getting killed because they did. Maybe along with their families.

A Mandarin-language CIA video appealing to informants, shared Wednesday on YouTube, was careful to note that “your safety and wellbeing is our foremost consideration.”

Until Trump gets in, and sells your identity back to your government in exchange for a hot dog.

I wonder how many of the toxic “left” accounts in the study were ones who also happened to show a suspicious pattern of echoing Russian-friendly or not-voting-for-Democrats-friendly talking points.

Certainly natural home-grown political toxicity is, as it’s always been, a feature of anyone on the internet who’s talking about politics, right or left. But I’ve absolutely noticed on Lemmy that the same users who are incredibly toxic about their approach to anyone who disagrees with them, also tend to sometimes have other anomalous funny ideas.