Bad decision making at Mozilla is in part responsible for the drop. They keep changing the interface and removing existing interface configurations (e.g. the new tab page configuration cannot simply display what you want it to; I also want the old tab bar). Mozilla could be more successful in keeping people if they focus on stability instead of interface changes. Youtube videos start stuttering at random and the sounds disappears completely when I use duolingo. I need to restart firefox at least 5 times a day.

It's worse than that. You're not turning a page. They printed half of it in black and the other half in blue and then claim you should have only read what is written in black because that's what they do.

It's a fact that most non-technical people don't understand what view source is. My parents or sister for sure have no idea. The analogy is necessary to communicate with people like that.

I had the impression M$ lost in court several years ago and were forced to ask users to choose a browser instead of pushing their shit.

What happens when the photos get leaked from StopNCII.org?

That seems like a first step in a tech arms race that could end with some government taking control over general AI and using it to subdue other countries

Everybody is talking about rowhammer as an attack. Is there any reason to believe this cannot happen by accident (thus leading to data corruption)?

I disagree. The network effect is very strong.

only 3 out of 4?

"less useful to know about" I disagree. Everybody should have some knowledge about digital security. I thought children disliking math is a western phenomenon. Is it the same, for example, in asian cultures? There are ways to make math more fun to learn (if education systems are willing to adapt and teachers care).

If algorithm they are using to determine what should be blocked gives a simple yes/no response, this is simply inadequate for the task. If they use supervised learning, training data could be annotated with what part of the content policy is violated. Unless their training data is based on somebody's gut feeling, they should in principle be able to determine what is being violated by a channel. When in doubt, a channel can be flagged to be reviewed by a human (if the channel is important enough). Otherwise, it's much better to have false negatives than false positives. False negatives will eventually be fixed as more people complain and without needing to write on twitter to get the attention of the humans in youtube. I don't know exactly what they are doing, but they have people whose entire job is dealing with this problem and they should be able to do better (I spent only 10 minutes thinking about it).

Futurology has been mostly clickbait and bs for a while. This "protest" is sufficient reason for me to leave it altogether.

The title is misleading. kot-begemot-uk summarizes everything pretty well

Judging by the results, one could form a hypothesis that speaking english is a major advantage to survival.

What's the benefit of allowing an AI to be considered inventor for a patent? So that the AI is more motivated to make more inventions in the future? This doesn't make sense and looks like a legal loophole.