tal

You can mount an ISO image through the BMC and boot the server from there.

I mean, you can typically do a TFTP boot on the systems that I’ve seen that have serial-capable BIOS setups.

considers

Heck, I think that all of my modern personal non-serial-port-BIOS systems can do that, even.

Baseboard management controllers are miniature computers that are embedded into the motherboards of virtually every enterprise server. The microcontrollers, typically abbreviated as BMCs, run with their own operating system firmware, network stack, and IP address. Administrators rely on them to monitor the physical status of large fleets of servers and to perform a variety of tasks, including rebooting machines, installing updates, and even reinstalling operating systems. BMCs provide what’s known as “lights out” and “out-of-band” management because they work even when servers they’re attached to are turned off or are unresponsive.

Hmm. I don’t really understand what the major benefit is to having an integrated management system over just providing a BIOS that can expose its interface on a serial port and having a serial device server attached to the thing, which then means that you aren’t tied to a particular management system. I’d also kinda think that it’d be cheaper to use a serial device server if you have a ton of servers in a rack that need to be managed, since one serial port server can handle all of them. Use a smart PMU for power control.

Maybe interfaces are more standardized or something, for automated tasks?

Eh, this one’s fixable, because you only have a few sources of legit imagery, and it’s all contemporary and digital. Just have Google Earth render the image server-side and add a cryptographic signature to any legit satellite images that someone saves.

Honestly, GNU/Linux isn’t better today from a user standpoint. I mean, Linux has the technical subsystems to sandbox software, yes — like, the move to Wayland makes isolating graphical software possible, where X11 imposed major limitations on one’s ability to do that. Various containerizing systems exist. But the typical software package distributed via Steam, say, in 2026, can see the filesystem as a whole.

Like, I can hack something up with something like firejail using that functionality that may-or-may-not work for a given game on a given system, but that’s not a real, universal solution for users. My mother is not going to do that.

Mobile OSes at least isolate most apps from most of the filesystem and some sensors, and provide some level of reasonably-accessible transparency into what apps can do. They still have their own issues that annoy me, like software normally being able to phone home with what information it does get. But that’s still way more isolation than a user who is playing a video game on GNU/Linux is getting.

Global, always-generally-available connectivity brought us a lot of nice things, but it’s also come with a lot of not-so-great things, and some of the not-so-great things is a loss of privacy and a loss of control over the things that we own. :-(

I mean, that’s really true of products in general.

I think that there are some real consumer law issues to answer where someone buys a product and then a manufacturer can control the product subsequent to the purchase. Like, say I buy a BMW with the expectation that I am not going to be seeing ads, based on the way it acts at the time that I bought it. At that point, I’m locked into owning that BMW. Then the manufacturer degrades my experience, maybe to the point where I no longer consider the vehicle to be the one I wanted.

Like, what is my remedy here? If the vehicle has OTA updates, I can never know what it’s going to be doing down the line, so I can’t make an informed decision. Having informed consumers is one of the basic requirements for markets to be efficient.

Maybe I could say “Well, that’s just a case of let the buyer beware. We’ll rely on reputation…some auto manufacturers will piss off consumers and those will be avoided”. But…if someone doesn’t buy cars that frequently, that’s a long iteration cycle.

I’m actually more concerned about video games. Like, if I’m a publisher, once I sell a video game, I don’t normally have more revenue coming in from it. But if I do have the ability to update that video game down the line. And I can update it to go harvesting user data, which turns it into a revenue stream. A lot of video game publishers go defunct on a regular basis or sell IP or whatnot, so there, even the “rely on reputation” route is more questionable.

It’s absolutely worth requesting your data to be deleted regardless of where you live by claiming to live in California…

I do think it’s possible that some of them won’t delete your data upon request, failing to comply with the law

That’s not failure to comply with the law if you don’t actually live in California.

People have a legal location of residence in the US.

If you want to establish legal residency in a state, you’re typically looking at doing something like spending the majority of the year there.

EDIT: I’d also be pretty confident, though the article doesn’t talk about it, that California won’t normally have jurisdiction for California state law unless a data broker is doing business in California, and that that will also be a constraining factor. Like, if a data broker company is only doing business in, say, Russia, the US will not enforce rulings against them regarding California state law in general.

They can’t make competitive dedicated memory in a near-term timeframe — the kind of timeframe that would matter, if they are to get something in during the shortage — and Cook has explicitly said that Apple will not try.

However, they do make their own CPUs, and they could maybe make a large-onboard-cache version, kind of akin to what AMD has been doing. That’d make the CPU more amenable to use of slower memory for most applications, which might open up more memory options.

“Technology, for a long time, has been promising people that they’re going to work less and they’re going to have all this leisure," Altman said, adding that things have gone in this direction.

He explained that people now have more time for leisure and a higher quality of life than in previous points in history.

"But somehow we never get the promise of the 4-hour workweek on a mass scale in society. And I don’t expect AI to change that,” Altman said.

OpenAI CEO added that people’s expectations always go up, and they want more.

“We think of new things to do, to create for each other, to want for ourselves. People are very focused on how they’re doing relative to other people,” he said.

Altman then spoke about the competition that drives the economy and people’s desire to be useful to others and to create.

Altman’s predictions are in stark contrast to the vision of Elon Musk, CEO of SpaceX, the company behind the chatbot Grok.

In a recent interview with the Economist, Musk envisioned a future in which jobs are optional, and money doesn’t matter, but he didn’t provide much detail on the transition to this level of abundance.

Yeah, I think that Altman is probably closer to being right than Musk is, at least for some time to come. If you’re talking about the kind of systems that OpenAI and xAI are working on, then certainly.

A lot of people — on the Threadiverse I see plenty posting — are really worried about permanent expanding structural unemployment. Something like “there will be no jobs” or something like that. And, okay, it’s not immediately obvious why that wouldn’t happen. And there have been some very prestigious people who have predicted otherwise.

npr.org/…/keynes-predicted-we-would-be-working-15…

The economist John Maynard Keynes once wrote an essay titled “Economic Possibilities For Our Grandchildren.” It was 1930. And in the essay, he made a startling prediction. Keynes figured that by the time his children had grown up, basically now, people might be working just 15 hours a week. David Kestenbaum with our PLANET MONEY team tried to sort out where Keynes went wrong.

To date, what has consistently always happened over time is that as our productive capacity increases, someone finds some way to create a higher standard of living using that productive capacity. That higher standard of living has always been more-appealing than the leisure time. If we wanted the standard of living that was present in 1900, say, we could work a lot less and just accept that standard of living. What happened instead of increasing leisure was increasing standard of living. Larger houses, air conditioning, cars, video games, elaborate movies, out-of-season fruits and vegetables, all that.

It’s possible that that phenomenon will end, but I’d bet against it in the near term. We’re really good at finding new things that people would like.

Now, I do expect that there will come a point where you can produce an AGI that is truly superhuman in the sense that it’s more-capable in every significant way than a human. I do not think that that is going to happen in a year or ten years, but it might not be that far down the road. And then, unless you see demand specifically for something made by humans — think of the way that there is demand for, say, expensive handmade fountain pens, something like Veblen goods — we’re going to have to rejigger things for a post-scarcity society.

en.wikipedia.org/wiki/Post-scarcity

Post-scarcity is a theoretical economic situation in which most goods can be produced in great abundance with minimal human labor, so that they become available to all very cheaply or even freely.[1][2]

Post-scarcity does not mean that scarcity has been eliminated for all goods and services. Instead it means that all people can easily have their basic survival needs met along with some significant proportion of their desires for goods and services.[3] Writers on the topic often emphasize that some commodities will remain scarce in a post-scarcity society.[4][5][6][7]

But I would be pretty confident that ChatGPT isn’t getting humanity to a post-scarcity society in the near future.

And I think that there are a whole host of other things that would change in human society in such an environment that I’m not sure that structural unemployment would be the first thing I’d think about. Like…what does the military balance of power look like in a world where an individual can hypothetically manufacture robots that manufacture factories that manufacture military drones, say? I mean, it’s a awfully different world in a lot of ways.

I hope it means that AMD is going to be more competitive on the parallel compute front. That’d put pressure on Nvidia to bring their prices down.

Righr now, Nvidia is able to charge an awful lot, because there’s little by way of alternatives.

Huh. That’s an interesting other aspect that I hadn’t thought of, where people might inadvertently make themselves hard to find with standard survival gear.

I think that this was just ordinary old cloth blankets that didn’t rely on reflection of radiation, but just had low thermal conductivity. It was talking about people sniping and such, sticking something above a foxhole. I can’t remember which military conflict I was reading about, but I’m pretty sure that it predated the Russo-Ukrainian War, since I remember reading about it early on, how this was going to be a conflict with a lot of drones all over looking into the non-visible-light spectrum, and how it was going to be a lot harder to hide from sensors now.

I distinctly remember reading about past use of of blankets to hide from infrared drone cameras.

By the popularity of type inference in variable declarations, a feature that’s included in most languages now: var in C# and Java, := in Go, auto in C++, and so on.

I don’t really like type inference, because I think that it makes it more-obnoxious to read code. That being said, I will concede that if you want a lot of static type information — and compilers can do things with more information — it does make for more concise code.

EDIT: Hmm. I was trying to think what would make it better. Maybe I’d be more okay with it if all compilers for languages that did this had a trivial way to just run the type inference step and output code with inferred types present.

There are so many surfaces in this world that suffer from wasted potential, are not yet displaying advertisements.

People who don’t like codeberg can just go somewhere else (even back to github) and/or found/support an alternative that better suits their needs.

searches

gitgud.io provides free GitLab hosting for open source projects.

gitlab.com has a free tier for personal/open source stuff, though looks like it has a 5 user limit on said free tier.

Gitea can be self-hosted. It doesn’t look like they provide free hosting themselves, though, just paid. Apparently it’s considerably less resource-intensive to host than GitLab.

GNU Savannah is available if one is writing projects under a specific set of licenses (the GPL family).

Investment provides resource allocation, which is not something that just happens on its own. That is, it says “this company or project is a worthwhile project to allocate capital to, and this is not”. The money here was also not invented out of ether, but rather lost by investors who shifted capital to SpaceX earlier, and gained by those who tried to shift capital away from SpaceX earlier; other investors came to agree with those shifting capital away. Those who do so earlier are rewarded; this encourages earlier efficient movement of capital.

Resource allocation is a necessary part of economic activity; any economic system will require such a thing to be done. You could have your government do it instead of investors, as in a command economy. Countries generally don’t do that, because it has run into more serious problems of misallocation.