Obviously. LLM-type AI is the ultimate enshittification tool.

Indeed. Why is this question even asked? Somebody really does not get it ...

Of course, only for the unwashed masses. The Epstein class and the other 1%ers will be just fine.

That is because you do not understand how this game works.

I have no problem with some (!) QC research being done. I have a massive problem with large amounts of money being spent on it and the proponents lying, lying and lying some more.

The "serious evidence" is very, very clear and does not need any more examination. Those that claim that this stuff will deliver an useful QC anytime soon are delusional or lying. Period.

Are you an idiot? I did cover the only case that certification is required. It is NOT universally required. Incidentally, the FSF may well go after getting that.

Or rather obviously, running AI agents productively is a _really_ bad idea at this time. But a quite large part of the human race has no active intelligence and needs to get hurt before they stop doing the stupid thing.

Which is completely untrue. The problem is the people writing unreadable shell-scripts (or any unreadable code, it can be done in _all_ languages). It is a poor craftsperson that blames their tools...

The thing is, most coders and other IT people stay in their narrow area and do not look out. Hence you get coders with no clue about system administration, networking and security. You get "security experts" that cannot code or do system administration or anything really (and are hence basically worthless). You get sysadmins that cannot do shell scripting (or any coding), which is completely pathetic.

This is a result from the immaturity of the IT field: No qualification requirements, no liability, no long-term technological stability and hence no real standard approaches and anybody can claim to be an expert on anything. And hence a lot of things are brittle as fuck, hard to use, hard to maintain, insecure and unreliable.

Probably. One reason why they are soooo far behind the rest of the west in many things, yet are convinced they are ahead. Essentially mass-delusion, since the information is readily available today.

It is actually both. The thing is that they can do a clean-room implementation based on the spec legally. But for that they need the spec. They are, again, allowed to reverse engineer binary code, _just_ to get the spec. They are not allowed to get anything else from the binary code.

Also, this may be done under EU laws, which are a lot more tolerant to making drivers and fixing bugs for hardware you own. They may run into RF restriction problems though, if those restrictions are in the driver and not the hardware.

Not necessarily. Questions of motivation, mental issues, malicious behavior, learning ability, etc. may kill that.

Same. This basically crap never works out and there are always a ton of not very smart people cheering for it.

The "not really AGI, but can navigate the world somewhat incompetently" kind? That one would be a huge, huge success actually, for quite a few areas, for example robotics. Friend of mine, a professor of robotics, told me that the "navigation capabilities of an ant would be a massive, massive improvement". It is unclear whether AI will ever amount to this much though. (No, Physicalism is not predicting that. They are using circular reasoning, like most religious and quasi-religious fuckups...)

Indeed. A lot of people cannot accept reality when it collides with their hallucinations though. Hence the low-quality attacks on Gutman's statements.