palordrolap

not sure what they’re trying to do here

Maximise profits and minimise losses. My guess is that someone important at Microsoft thinks that this will do just that, and if not that, will make them, personally, a lot of money. That person has no-one who will dare challenge their authority and so we go down this road.

They (that individual or Microsoft as a whole) almost certainly have a stake in the companies that provide newer hardware, and if they didn’t before this decision, they will have by now.

It theoretically makes Micosoft’s job easier too. A huge chunk of backwards compatibility maintenance goes out of the window, if you’ll pardon the pun.

“Oh you have 5 year old hardware? We don’t support that.”

Sounds fairly similar to Apple’s business model if you think about it that way.

There’s a reason the word “multifaceted” gets applied to people sometimes. There are different faces we present in different scenarios.

Sometimes people will be one way with family, another at work and another way again when with friends. There are funny stories (or webcomics) of people ending up in a group that comprises a family member and two people they know from respective second and third groups and the humour is that they have no idea how to act in a way that won’t confuse at least one of the people they’re with, including themselves.

As for relating to your situation, there was one online content creator who went by a name that sounded like it was based on a real, given name and not something made up. It was a bit of a surprise when I learned that wasn’t anything to do with their actual name. They also posted things that were out of the mould I’d mentally put them in on another social media site.

(Being vague because I don’t want to dox or disparage.)

Thankfully, those particular revelations affected no-one but me, which might be a way for you to look at the situation with the person you’re talking about.

Current AIs often suffer from what’s sometimes called “the strawberry problem” and can be easily confused into doing mathematics incorrectly. There’s also the human element of “if it comes out of a computer it must be right”, forgetting the “garbage in = garbage out” principle.

The strawberry problem is a colourful name for the fact that LLMs turn sentences into something else internally and can’t then go back and re-examine the input to make checks and comparisons. Thus if you ask an LLM how many 'r’s are in the word “strawberry”, it often gives the wrong answer, unless it has been explicitly told what the right answer is. And now you have no idea what else it has and hasn’t been told is right and wrong.

As for mathematics, they have to be explicitly programmed to be able to use a proper calculator if they’re to do mathematics correctly. Otherwise you’ll get something that looks good enough to fool someone who doesn’t know any better.

LLMs are basically lossy compressions of knowledge. At a high level of abstraction, the creation of an LLM is fairly similar to how a raw image is turned into a JPEG*. There’s a necessary, deliberate bottleneck in the creation process that keeps the size down, and that’s going to show up in the output if you look closely.

Using the output of an LLM is a bit like editing the JPEG rather than the raw image. Some of the things you do will invariably enhance those artefacts.

For a JPEG that’ll do nothing worse than make an image “deep-fried” or otherwise ugly. Put an LLM in charge of people’s lives and it’ll do the same to them.

* JPEG has a lossless encoding variant, but that’s not the right analogy here.

With the right (or maybe that should be “wrong”) people out of the way*, many more people will be fed than would otherwise not be. Short term goals versus long term goals.

Also, stale baguettes, while technically food, are not the most nutritious of foods. Employing the baguettes in other ways may result in better nutrition.

* I should make plain** that “out of the way” does not necessarily mean the most extreme measure. It can, but it doesn’t have to.

** Somewhat like a stale baguette.

Web-rings were also a thing in the mid-to-late '90s before search engines really took off.

Basically you’d put a banner on your site that listed a site “before” yours in the ring, the one “after” and maybe a “random” link as well.

Wikipedia: https://en.wikipedia.org/wiki/Webring

About 30 hours.

I had to go nil-by-mouth for 12 hours before an operation to repair a fairly serious injury and they kept pushing the surgery back and back and back. Higher priority cases were keeping the surgeon. It wasn’t like I was low priority either, but my injury was stable and not immediately life threatening.

Did I mention I’d also lost blood? That made for a force multiplier.

In the end, they admitted defeat - the surgeon had worked too long anyway - let me eat something and rescheduled my surgery for the following day.

Let me tell you, that was the best chicken I ever ate.

I can imagine that works well. Once.

Typing in a box on a website provided by someone else can be done by practically anyone, even the boss themselves at a pinch. Heck, some of them love doing that.

Maintaining a server (or a service on a server) requires someone with more skills, which generally costs more money.

Yeah, but then they have to pay someone to maintain it. And low though that cost may be, it’s precisely the sort of thing that C-levels like to cut when they feel like pretending that they actually do something.

It’s a hard thing for a C-level’s ego to set something up that they know someone else is going to cut later, or worse that they might be the one to “have to” cut it.

Very little of what is your body now will remain in a few years. That’s less terrifying than it sounds because we replace bits of ourselves on a constant basis. With every breath we lose carbon that may have been in us for years. Every bathroom visit contains not just food remnants but little bits of ourselves that have already been replaced and broken down.

About the only things that don’t get some level of replacement are our teeth and the floaters in our eyes.

As slurs go it’s pretty low tier. A bit like the French calling the English “rosbifs”, if they even still do that.

They should be using tablets. Wax tablets. It was good enough for my great-to-the-nth-grandfather, so it’s good enough for kids today.

Do you remember what it was like before you were conceived? Like that.

Geolocation is done by IP address if nothing else is available. It’s not exact, but if, for example, you use a US-based internet provider, all their customer IP exit points will already be marked as US-based, and any access from one of those to a Google service will get the new name.

And it may yet swing back the other way.

Twenty or so years ago, there was a brief period when going full AMD (or AMD+ATI as it was back then; AMD hadn’t bought ATI yet) made sense, and then the better part of a decade later, Intel+NVIDIA was the better choice.

And now I have a full AMD PC again.

Intel are really going to have to turn things around in my eyes if they want it to swing back, though. I really do not like the idea of a CPU hypervisor being a fully fledged OS that I have no access to.

If you like a bit of pop pseudoscience, you might like* the idea that today (or this time last week), is Blue Monday: https://en.wikipedia.org/wiki/Blue_Monday_(date)

* Take this word as you see fit.

It’s a rich text box with a few controls to enable or disable those features at certain points in the text. The whole thing was entirely built from components used elsewhere in the OS, or at least the earlier versions were. One competent employee could manage it in an afternoon; a week at the outside. If Microsoft has let it get so ridiculously bloated that it’s now unmanageable by one person, that’s on them.

I never said confusion

People expecting Word capabilities

Now, why would they expect that?

See also: Java and JavaScript.

They could have just renamed it. Wordpad’s Win3.11 predecessor was called “Write”, for example, so that name could have been revived.

For a long time, write.exe still existed and all it did was launch Wordpad, so they’d only have to reverse that.

They could also have chosen another name entirely. Or, since they’ve recently added a bunch of unnecessary crap to Notepad, they might as well have merged the two.

“Confusion” is merely an excuse.

Very much this. I can see one potential future Microsoft product being something that is to be installed on a thin client PC sold to consumers for cheap. It will run not much more than a browser in which all apps will load from Microsoft servers, and all storage will be on the Microsoft cloud. And if you miss a monthly payment they’ll basically hold all your files for ransom until you start paying again.

I can practically hear the Microsoft execs making some very unsavoury noises about that idea.

As for (admittedly somewhat weak) proof they’re headed in this direction: Wordpad is a useful small program that would easily fit onto a thin client and there’d be room for documents created by it on the limited storage available. It has to have some storage for browser cache after all.

Wordpad was recently cancelled, and users urged to use Word instead. Which is not free of (further) cost like Wordpad was.

I guess I do. I put the computer (a desktop) into suspend most nights so that it’s pretty much up and running as soon as I turn it on the next day.

Even so, rebooting doesn’t take that long. 30 seconds tops. Definitely not enough time to visit the bathroom or make a hot drink.

But the advantages to suspend are that it’s quick and all my programs are as I left them. A reboot undoes most of that.

Yes, hibernating is also an option to keep open programs, but why do that when it can be quicker?

My only real concern with putting the machine into suspend is if there’s a power cut and things end up in a weird state or I lose work because programs weren’t closed properly, but then, that could happen at any point when I’m using it too.