Thanks for making it clear that you are not interested in the facts of the matter. How pathetic.

Yep. Commercial education is something that does not work.

Sorry, but "multiplication tables" and doing basic algebra in your head is not Math. It is bullshit make-work. You can carry a calculator (most of us do), a slide-rule or pen and paper. What counts is whether you know what multiplication does and how it works, not some rote memorization. I say that as a PhD-level engineer.

Indeed. When everybody graduates, the degree becomes a complete joke. And that is really bad for society, with long-term effects.

The reasons are typically too many people incapable to admit things are broken (and here: admitting that their kids may not be so smart). Denial kills everything because it kills feedback loops.

Massive. As in "let's copy anything we can".

And no, data processing to calculate LLM calibration is not "training" in the same sense as a human does it. Courts and the law understand that, even if idiots like you do not.

You think wrongly. No experience with data analysis?

Here is a starting-point: https://matthewdwhite.medium.c...
Here is another: https://ml-site.cdn-apple.com/...

On the side that claims LLMs can reason, it is all marketing materials only, i.e. bullshit.

I see some stupid people have mod-points. Here, waste some more. I dare you.

so as long as there have been android tablets, I've had a 7-8" one that is my Book. fits in the back pocket, holds a hundred pounds of books and a bunch of manga too. Right now I have a Samsung Tab2 S2 that's showing its age a little in terms of battery life, but I can't see myself going back to analog books, barring some kind of apocalypse.

Naa, that would have been competent. AI people do not do competent. The ones that did renamed their occupation to other things, like cognitive systems, automated reasoning, robotics, etc. The ones that still claim to be doing "AI" are the hacks that are in it for the money.

The proper way to do this is 1) fake it and 2) when queried, lie about it. I mean, this has been the traditional approach in all things AI and at least the LLM pushers know how to do it. I would have thought that Russians, off all people, understand this approach in a more general way. Apparently not. Some people will probably get an extensive "vacation" sponsored by the state now.

Ah, yes. "Code slower with LLM assistance". Also probably code worse. Good luck.

Unless you think FFmpeg has tons of legitimate security problems, the volume implies the slop being in there.

Have a look at the numerous times Azure got hacked and the utterly ridiculous security failures by Microsoft in there. Then their clueless bumbling with Windows suddenly looks like a minor issue.

I mean, it is so bad that they now try to bribe people. That is the clearest admission of failure they could give without scrapping the whole mess...