The article gives a very strong impression that this is not for the benefit of users at all, that this is for the benefit of the companies, so whether it helps visually impaired users is not relevant.

According to a 2018 trial with Yahoo users, Verizon said that after adding verified logos next to inbox emails, they saw a 10% increase in customer engagements, as users tended to click on emails with a logo more often, driving traffic to companies which tested the technology.

"Last I checked" -- What did your check tell you about "Twitter Communications India Private Limited", registered in India since 2013?

If the police do suspect you're using such a hidden container, what happens then if you are unable to convince them you aren't?

What he did was horrible, but 5 years for watching someone on camera seems excessive.

To put this into perspective, this is less than a day per offence, or just over a week per customer account violated, though.

No one was physically hurt, no money was stolen, he’s essentially a peeping tom.

The summary states that he persuaded at least one customer to have cameras installed that she had not wanted, which presumably she paid money for.

A normal warranted search is breaking & entering, but the warrant makes the specific case of breaking & entering legal. Likewise, it's very well possible that this is fraud, but fraud that is signed off by a judge beforehand is still okay.

In their WSL2 FAQ, https://docs.microsoft.com/en-..., Microsoft specifically say that WSL1 is not abandoned though.

This looks pretty nice, but why is this WSL2-only? The MS blogpost doesn't say, and it's not obvious what this needs to do other than provide services that Linux applications connect to, which WSL1 also supports. Any thoughts?

It's pretty obvious that there are differences between people, it has been illegal to use those to discriminate if that means you're actually discriminating based on a protected characteristic for a long time, and that isn't full-blown 1984, it actually becomes quite obvious if you start to look at some concrete scenarios. A trivial example is parental leave. This by itself is not a protected characteristic, but if you have any policy based on how much parental leave people take, then that policy is almost certainly going to result in gender-based discrimination. Another example are certain clothing choices. This by itself is not a protected characteristic, but if you have policy based on the presence or absence of headwear, then that policy is almost certainly going to result in religious discrimination. There are plenty of other examples you can think of. People know that policies shouldn't be based on that unless they have a very good reason, but AIs don't. If you only filter out explicitly protected characteristics, you would still be feeding indirect indicators of protected characteristics into the model and you would still end up with AI that models any historic biases.

You answered your own question: it was legal for the FBI to do this because they obtained permission from a court to do so. There is no sense in charging the FBI for doing something that a court explicitly allowed them to do.

By "referring to someone in a gender-neutral manner" I thought you meant "referring to someone in a way that does not make any statement about gender", but I think you are talking about people who use "they" to refer to people who themselves use "he" or "she", is that right? Can I take a slightly different view on that? When the people who do that do consistently use "he" or "she" for the vast majority of people, only using "they" for people who do not sufficiently fit gender stereotypes, it becomes quite obvious that "they" is not used in a gender-neutral manner in the first place, that "they" is actually used *to* make a statement about gender rather than to avoid making one. And that, of course, is misgendering. On the other hand, the sentence "LenKagetsu wrote a comment about gender", as an example, does manage to refer to you in a way that makes no statement about gender. That is not misgendering, is it?

Some people prefer not to be called black. Without more context, I would just take this to mean that black is the term Gebru and Curley themselves chose to use.

Yes, got it already from your previous message. Freedom of speech is the only right that matters to you, and if that right is threatened in any way, that's heresy and those people who threaten it should be forcibly silenced. If other rights are threatened, that's okay, because those threats are freedom of speech.

Framing freedom of speech in such a way that anyone who disagrees with you does not get to enjoy it. Got it.

People who are against the principals of free speech need to be aggressively stamped out.

That is pretty clearly denying them their freedom of speech.

It appears that the full results of a web service were included in the HTML, not the passport number specifically. Where the problem was is not clear, it could be that that web service was supposed to only return the data the page was supposed to have access to, in which case including the full results was not inherently an error. It could even be that initially, the web service did return only the data the page was supposed to have access to, but it was then changed to return more. The result was obviously bad but it's not necessarily a really bad coding practice anywhere, it could just be poor communication between different teams (which, I suppose, is still a bad non-coding practice).