No, they're a lie.

The infographic is propaganda, and answers zero questions about how the factors which make up each group are actually calculated.

I have been watching Experian. There hasn't been anything unrelated to me appearing on my credit report, but my score has gone down as I have behaved faithfully and paid off my debts on schedule. Anyone who believes what FICO claims about scoring is a fool, and anyone who then goes on to repeat their propaganda for them is also a tool — and not the sharpest one in the shed.

This is very true, and perhaps continuing to close the gap is their strategy for defeating CUDA, since they sure aren't putting enough effort in to do it with ROCm. I'd love to see it, I just don't expect it to work. Happy to be proven wrong, though.

My score was 790 a year ago. I faithfully have made my student loan payments, paid off a dental loan, and increased my available credit, and my score is now about 720. Credit scores are a scam and basing anything but applications for credit on them should be a felony.

That's true, but you also can't run old Windows on new hardware, and getting graphics performance out of Windows running in a VM is only even vaguely possible with vmware. Virtualbox's video drivers for Windows are shit, and QEMU/KVM's are even worse. Hyper-V's aren't great either. So you have to keep an old machine around for old windows if you want to run old windows games on windows.

AMD has been outselling Intel in the DC for what, a year now or more?

I don't think you know what "ensuring" means

You don't need lies to turn children away from religion. History and facts will do.

We will have sham elections, like the last one, where there were Russian bomb threats.

Usually SSDs get slower as they get more full. Most SSDs do not have much additional commit space. The amounts have been tested for, we discussed it here not long ago IIRC.

Right now "society" is making it hard to even be a student. Did you know it's harder for a student to qualify for SNAP than a shiftless layabout? If they're half-time or more (usually 7.5 units) they have to meet an exemption to get food aid, while we give it to people who simply refuse to work. (That does not describe most SNAP recipients, but of course it is a percentage.)

Did you get into the bad crack this morning?

*used to compete, since Intel hasn't had any parts competitive with AMD for years now.

Nvidia is also bringing out their own ARM-based servers, so far the point is to run their GPGPUs cheaper than with amd64 but when the AI bubble collapses they may well have to pivot in that direction to keep up DC sales.

Maybe, except that their AI chips aren't as good at running LLMs as Nvidia's CUDA cards.

The thing is that places that historically didn't have to worry about this, now have to worry about this.

If they want it preconditioned? Yes, welcome to 2025, they can install the app on their phone. Or they use the 'remote climate start' option on the keyfob. Or they shoot you a quick text asking you to hit the button in your app.

You keep trying to paint these advancements in convenience and comfort as terrible burdens, and it's weird.

Can't really embarrass anyone who supports a holocaust because they're racist against Palestinians.

Yes, I have tried to encourage the genAIs to ask questions a number of times in various ways, but with no significant success that I can recall.

I can definitely recall the results of my last paired experiments. I prepared a short software specification and gave that text to DeepSeek and ChatGPT.

The DeepSeek result was better in terms of how I described the desired appearance, but one of the four results was wrong. DeepSeek had clearly misunderstood that part of the problem and did NOT ask for clarification. But I admit that I didn't notice the error until I compared it with the results from ChatGPT's second attempt.

ChatGPT failed completely in its first attempt, but its second attempt apparently threw out the appearance parts and produced the right numbers. When I asked ChatGPT about the failure of the first version, its first (verbose) response was useless, but a later response sounded like it was trying to do the right thing--but I still have no idea why it failed so badly. The second version from ChatGPT only required about 70 lines of code, which I then annotated extensively for an acceptable result.

My main problem with the verbosity is that I wind up skimming thought lots of irrelevant stuff hoping to find something that is significant...

Well I can report that just telling the genAIs to be less verbose doesn't help, though I recall DeepSeek promising to try.