I had my clanker (Qwen3.8-abliterated) try for over 10h to weaponize some of the best looking PoC with no results except crashing workers.
ASLR was the first thing that blocked it, then it tried with it disabled for that process. It did manage to reliably crash a php-fpm or apache2 process with bad input.

So yes there is a problem, bad stuff is happening in the memory management and there are bugs.

ASLR is partly to thank for making exploiting this significantly harder.
There's also a lot of things that might be tied to the specific version of glibc and other hardening aspects.

Ask it what happened in Tiananmen Square in 1989...

Of course, being open weights, abliterated is a wonderful word to see after a model name.

So the 4 countries which traditionally hampers these efforts aren't going to be involved.

So there might actually be something meaningful after all.

They are focusing on innovating. Sure some numbers are probably fudged / inflated. But overall, they're kicking the western world arses when it comes to PV, EV and battery tech.

Like playing slot machines, sticking money into it and expecting a big pay back. Most of the time you have a small loss (time wasted on a bad generation, cost of tokens, etc), sometimes you have a middling success (some small productivity gain), few times you hit a jackpot and actually get something really good, but these are really rare. They do however get you hooked in for more very well.

I do use AI at work (Claude) to some success, but I don't think that the gain is really going to cover the additional costs of using it. Lots of time its incomplete or wrong in a small way, sometimes its even worse than an outright failure. Yeah boilerplate is quick, but often devs have their own tools to do it.

Personally, I've been using Qwen Code on a very old personal project I gave up back in 2000. Initially things were going well, major language updates, some rewrites etc, tackling obvious bugs.
But small bugs have taken multiple days of testing, prompting, testing, etc. Context is fleeting at best, the memory file is often ignored a few prompts in. Test generation has been off the mark lately as its clearly not used to BBS door games.

It measures economic activity. Not quality of life. Not quality of the environment. Just how much value is extracted per population on average. A baby will have the same per capita GDP as whoever is the richest person in that country.

It does not reflect anything on the quality of life, nor on the actual wealth of the population.

A partly trained intern would learn and get better. An AI model keeps making the same mistakes, again and again and again.

There's quite a few signs that AI made a piece of music, there's a general grittiness to it, the voices are very similar no matter the band.

Draconian, Nanowar of Steel, Sirenia, Can Bardd, Be'Lakor, Cradle of Filth, Forest of Shadows, Babymetal, Moonspell, Amorphis, Dark Tranquility, ...

At least its not part of a captcha system...

Looking at the picture in the article, the guardrails are still there, despite the bridge being washed out. This is also a contributing factor, giving the illusion that there is a proper road in place.

Whenever I'm stuck in traffic and forgot to put Waze on.
We have a local station that is solely traffic reports.

I'm not even sure what the issue is...
I have a Nissan Leaf (EV) and AM works fine.

This is more akin to building a house and not putting a lock on the front door.
And this isn't even the first time a car brand got attacked similarly.

Nissan had the same issues years ago. It took months between the discovery of the vulnerability and for Nissan to fix it.

I use gentoo on my home desktop. Its a major pain to update it, and you have to keep doing it, otherwise you have even bigger pains.