Plot twist: North Korea deliberately leaked the phone as a trojan horse in order to spy the BBC from the inside.

Sounds like the government is giving big tech companies access to poor people's data for pennies.

AI is certainly a problematic technology, but what really is nightmarish is that such technology is the latest and most powerful lever in the hands of gargantuan tech companies that are devoid of any moral considerations and wield more power than any person or organization in the history of humanity has ever had.

Even more red tape and gatekeepers who get a say in who can communicate over the Internet and how hard it has to be. I remember when it was Microsoft who wanted to make internet protocols more complicated in order to get a competitive edge over the open source community; back then when their plans were exposed there was outrage. Nowadays Google and Apple basically are the internet, and they don't need to work in the shadows to subvert the protocols, because whatever it is that they decide for the day is automatically the "living standard".

Of course, by all means, but so long as it doesn't turn into an antiscientific witch hunt against people who do save lives. Lawyers also profit from lawsuits from patients, as does everyone who gets paid to solve other people's problems.

Everything is easy if you choose to ignore complexity.

How many of the projected new cancer cases involve patients with old age or pre-cancerous conditions that would have evolved into cancer anyway due to other factors? Statistics are hard to read.

I was a kid in the 80s and as far as I remember no one talked about Tron, or even knew that it existed. Maybe it had more mindshare among cinema enthusiasts than among its target audience.

A copy is a copy, I don't think you can argue that a certain kind of copy doesn't count for technical reasons.
Check the case of the Internet Archive, IIRC some time ago they lost a lawsuit over digital lending even though they had gone to great lengths to ensure that, at any given time, only a copy of the book that they were lending existed inside the memory of their servers. The judge wasn't convinced.

Copyright law gives an author a monopoly over his or her work. You can only copy that work either with the author's permission OR for a reason that qualifies as "fair use". I'm convinced that creating derivative works that compete commercially with those of the original author cannot qualify as fair use. Then again, IANAL.

Any work of art that the AI couldn't produce, hadn't the original artist's creations been copied in bulk to be used as an input.

Good luck demonstrating in court that copying the whole production of an artist in order to sell derived works of art is fair use.

I don't want law to be enforced on the weak and interpreted for the strong. No one is going to junk the copyright system because of AI: what lawmakers will do is to enshrine in law whatever it is that AI company need to do in order to get money and power, and screw the artists or intellectuals who will lose their livelihoods because of it.

Did they *copy* even one byte of a Studio Ghibli work into one of their computers' memory in order to train their model? If so, they did *copyright* infringement, because only Studio Ghibli have the right to decide who can do that copy and under what terms. I don't know how one could possibly consider this behaviour to fall into a "gray area". The case was considered clear-cut when big corporations were suing moms for millions because their children had downloaded a couple of songs over bittorrent.

Don't try rewriting history. Games were much easier to copy, and copying them was widespread, long before CDs even existed. If anything, CDs for some time made games harder to copy. There was no "generation shift" in the 2000s that forced publishers to introduce DRM. It's just that the technical and legal environment was ripe for introducing this anti-consumer measure.

There is a problem in general with the abbreviation of strings in user interfaces, when the length of the visual representation of a string exceeds the width of the user interface field that the programmer has designed to display it. This problem has always existed, as this 30-year old bug shows, but has become much more widespread today, when programmers don't design and test user interfaces directly but rather use tools and technologies that render the programmer's abstract description of the UI into a concrete presentation that is specific for the resolution, density and orientation of the screen that is currently displaying it. This is great in theory, but in practice what happens is that programmers only test the UI in the English language and on an expensive large screen, and then the rest of the world gets to choose between "document version 1" and "document version 2" by means of a combo box that shows two identical lines reading "document versio...". Often, with no scrollbars, no draggability, no tooltip, without anything that could let one distinguish between the two options. And sometimes even without the "..." telling that a shortening happened at all.