He did not write AI (uppercase A, uppercase I), he wrote Al (uppercase A, lowercase L), referring to someone by their first name.

Without paying, people face a usage limit on how much they are allowed to use it;

No, they don't. Read the license carefully: firstly, Clause 2 does not distinguish between non-commercial use and commercial use, despite the "Additional Commercial Terms" title, and secondly, the restriction is based on the monthly active users on the Llama 2 version release date.

This means that if you had 700 million monthly active users on the Llama 2 version release date, you are not permitted to use Llama 2, whether commercially or non-commercially, regardless of how many users you have at the moment you use Llama 2, unless you obtain a special license from Meta.

This also means that if you did not have 700 million monthly active users on the Llama 2 version release date, you are permitted to use Llama 2, whether commercially or non-commercially, regardless of how many users your organisation grows to. If, like almost everyone, you fall in that category, nothing in Clause 2 will ever limit you, and Llama 2 is free for commercial use for you.

No one claimed it was an open source license. It's not, and if someone did you'd be right to correct them, but no one did here. As for commercial use, we can absolutely use this, even commercially, without having to pay for it. That is how an ordinary person would interpret the phrase "free for commercial use". Likewise, your "not free for commercial use" implies that people who are permitted to use it non-commercially suddenly face restrictions when they want to change their use to a commercial one. That is simply not the case, there is no such restriction. If you're interpreting "free for commercial use" as something other than "free for commercial use", that's on you.

Read it yourself, https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fgithub.com%2Ffacebookres..., the only relevant clause is clause 2, which is a restriction that does not depend on whether the use is commercial. Llama is free for commercial use for anyone not subject to that restriction.

I know nobody likes to RTFA, but literally, it's right there: "Code Llama 70B is available in the same three versions as previously released Code Llama models, all free for research and commercial use:".

Don't forget Danese Cooper's statement at Debconf 2006 that the CDDL was based on the MPL in part because the MPL was not compatible with the GPL. Wikipedia has details along with a link to a recording of her saying just that. For context, as also written there, not everyone agrees with her on this.

The idea of sticking your finger in the dyke to prevent a flood comes from "The Little Dykeman", https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fbooks.google.co.uk%2Fboo..., and the child is regarded as a hero, as a saviour of the country, for keeping the sea out all night until an adult finally came by the next morning. That analogy does not support the argument you are trying to make.

You have a real problem with terminology. Subordinate, worker, underling, all these words mean what they mean, and they do NOT imply the boss yelling or berating or barking. If that's how you interpret them, then that's on you, not me.

Subordinate and underling do both have definitions that are inappropriate to use to refer to your employees, and in my own personal experience, those definitions are how the words are most commonly used nowadays. Your experience may be different. You're not wrong if you use the words the way you do, but I suspect you will be more likely to encounter other people who also do not understand what you mean.

Not sure that would apply in general. But it's not important: in this case, Dutch law clearly applies because the employee was employed through a Dutch branch of the organisation: "The resident of Diessen, Noord-Brabant, was hired by the the Rijswijk branch of Chetu Inc., a software development company headquartered in Miramar, Florida." A Dutch branch of an organisation has to abide by Dutch laws.

(and elements of the suit are still being litigated)

It would be useful to know which elements of the suit are decided already and which elements are still being litigated; it is hard to find up to date information about this. The most up to date news on this I can easily find is that "[a] skeptical Fourth Circuit panel signaled it may nix the music industry’s $1 billion copyright infringement award against Cox Communications Inc." (https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fnews.bloomberglaw.com%2Fip-law%2Fcox-appeal-probes-1-billion-music-copyright-loss-for-cracks)

On the other hand, if you were AWS or Azure, how would it look if this attack gets used successfully to leak sensitive data and you knew of the possibility of that and the mitigation to prevent it from happening beforehand, but chose not to use that?

If its not 'intended' than it is by definition neutral on the subject.

That's not how the law works and the rest of your message already somewhat disagrees with this. The keyword to search for is "disparate impact". Discrimination on something that is not itself protected but that has a "disparate impact" on a protected group, is still illegal in many places including California, regardless of intent. You're right that there is an exception when there is a legitimate reason for it, but the burden of proof for that lies with the employer (per the Civil Rights Act of 1991), I'm not convinced your example would qualify as a legitimate reason, and in the case we were originally talking about, board members, I'm convinced no legitimate reason could exist. Proving that such discrimination happened though, that's difficult.

If you have 10 companies where the board is made up of 9 men, how many of those companies do you expect picked the best 9 people for the job, and how many of those companies have something, intended or not, that prevents women from taking up the role?

It does not. The ruling singles Google out: "Google, ein Unternehmen, das bekanntermaßen Daten über seine Nutzer sammelt und das damit vom Nutzer empfundene individuelle Unwohlsein so erheblich, dass ein Schadensersatzanspruch gerechtfertigt ist". Translation by Google: "Google, a company that is known to collect data about its users and the individual discomfort felt by the user is so significant that a claim for damages is justified". This would not apply to CDNs in general.

The site literally does ask for your password, at https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fhaveibeenpwned.com%2FPas.... That's not an attack on Troy, that's a simple verifiable fact.