I'll assume you are being serious.
1. Not all AIs are equivalent to ChatGPT.
2, Mistaking something that isn't a vulnerability for a vulnerability is relatively low cost.
3. Finding one vulnerability that's real can be extremely important.

NOTE: It doesn't NEED to be perfect. If it's "good enough" then it's good enough to be useful. Things that aren't vulnerabilities are relatively cheap to check.

P.S.: You shouldn't have needed this explanation.

If you're hired specifically to develop a good AI, then you *should* push back against folks that want you do act in ways that would cause you to create an inferior one.

OTOH, for Meta to fail at this is devoutly to be wished for. I'd prefer almost anyone else.

A lot depends on how much you believe their explanation. I don't. In fact, I suspect the person making the explanation didn't know the reason, and either invented what they thought would sound good, or just read something someone else handed them.

Corporations don't have a "central mind" that knows all the things they are doing and why they do them. To get a reasoned answer takes a long time, and usually isn't what they want to deliver anyway.

Since China was already trying to get local chips used rather than NVIDIA, they'll probably be in favor of this move.

This needs an anti-trust lawsuit.

Using Microsoft default anything should be a felony.

Hopefully this will be enacted in Europe soon!

It is not just the artificial "intelligence" that is becoming enshitified. The impact on the real thing is every bit as serious.

Bitcoin does have some small intrinsic value. But it would be around the value of a ticket to the 1933 world's fair. Or an old newspaper with the headline "Dewey Wins".

Communism is not a workable system for more than Dunbar's number of people, and no country on earth uses it.
I really don't think it would work as an economic system, either, for the same reasons.

For groups smaller than Dunbar's number, that also have a charismatic leader, it can work quite well. But when that leader fails or retires, they tend to adopt a different system...or just fall apart.

That's not quite true. Occasionally some people in government do want to reduce the power of the government. For some reason they never end up making the decisions. This is because "power" is an instrumental goal that even an anarchist would desire if they wished to further their belief.

Every one dimensional metric oversimplifies things. But "fascism" is not well defined enough to use as a metric. And "statism" is the wrong term, if you're going to contrast against "individual freedom" the opposite pole should be "authoritarianism". E.g. many small communities traditionally didn't have any central government (i.e. no state), but they insisted on strict conformance to their rules via social pressure. (In that case the "authority" wouldn't be a person, but a set of social rules.)

Read my post again. You didn't understand it.

Last I checked Ruby execution was slow compared to Python. That, however, tells you where you shouldn't use it, not *that* you shouldn't use it. And Ruby can easily call C routines (with the usual caveats).

OTOH, in some task spaces, design in Ruby is fast compared to design in Python, and in almost all it's fast compared to design in C. (That said, I generally prefer to design in Python and then re-implement in C++.)

Whether it's serious or not depends on what you're doing. For me it fails only because I require Doxygen compatibility. (Mind you, I would rarely choose to use *only* ruby, but for some things it would be the superior choice.)

OTOH, Ruby is not a low level choice. It's a slightly higher level than Python. And I often design things in Python and then convert them to C++ (with, of course, minor rewrites).

So, "What do you mean by 'serious'?".

The real question is "Will they re-release the roadrunner cartoons?".