Nowadays it should be:
* AI generated smart TV in VR
or
* smart TV in AI generated VR
or
* smart TV in VR displaying AI generated content
or, even more exciting (contains more AI acronyms)
* AI generated movie about a person entering VR to watch smart TV (potentially a movie about AI generated movie about a person ...)

I've noticed that stupidity with units too. They also give millions of tons of CO2 in the report. For a reference 31 TWh is amount that is used by a whole highly developed country of about 5-6 mln people, e.g. Slovakia (26 TWh, 5.5 mln), Ireland (34 TWh, 5.4 mln) or Denmark (36 TWh, 6 mln).

How about if I create an "AI" that is trained on a single book (could be latest bestseller by pure coincidence), and then "recreate" the book, and sell it as a AI generated book. Would you consider that fair? If not, where is the limit in number of books you need to use in training in order to your "AI" to be considered fair?

There are over 400 world-wide working nuclear reactors, there are proven to work, and been improved for over 50 years. All of them cost in range of couple billions, and time to build one from the start is in range of 10 years. There is not a single working fusion reactor, and none is even close to be working continuously, and even dreaming to be used as a power source, not just a research device. We are not even at the level of Chicago Pile 1. But sure, there are some guys that will do that for 65 millions in 6 years. Thank you EU for dumping 35 millions in this cleaver project. Thank you for NOT giving this money to the real science, like ERC grants, where there is such a tough competition, and a lot of very good project do not get funding.

It is high time to do something about the publishers. They have free papers written with LaTeX templates - these are almost publish-ready documents. They ask referees to check the papers for free. Then they charge you $12690 (Nature) or $4000 (APS) for open access publishing. Or they charge governments and universities in millions of dollars for national access. What are the overheads? The costs of publishing (online) or little editing after the referees have done their job, seem to be negligible compared to the prices. The whole thing is run on impact factors and reputation.

I would like to see CERN starting it's own peer-reviewed journal. In a couple of years those vultures will all be shot down.

I must admit I use sci-hub from time to time. My university have some subscriptions with editors limited in time range. For example a paper from 1987 might not be available on-line, but we have, almost always, this journal in paper copy in a library. I'm lazy, so instead of going to the library and scanning the paper by myself, I simply use sci-hub. I don't feel like pirate doing this. Anyhow, if a paper was retracted, the paper copy stays in library (same as sci-hub). Unless they send some sort of legal requirement of destruction of all the copies, which I have never heard about in regards of academic journals.

Who gave the reference diagnosis then? The one 100% accurate?

In short: 50 patients were studied by real doctors in real hospitals and clinics, and they get a proper diagnosis. Whatever was written in the papers - short history of present illness, past medical history, and symptoms (e.g. temperature, pulse, skin description) - was given to other doctors and LLM. What is shows is that people, to get proper treatment, need direct contact patient with a doctor. This is what doctors are taught, and expected to do. LLM or online consultation will not replace that.

Sure, the atomic bomb is a terrible weapon. But we should also ask about families and survivors of about 20+ millions of japaneese war crimes commited with more traditional methods, such as swords, guns and fire. Those are equally terrible. And it was not like suddenly the bomb was just droped on some innocent country...

It's a World War I situation. What to do if an artillery barrage doesn't work? Both sides commanders had only one answer: bring more artillery.

I have a 13-years old laptop (HP 2540p) which I still use. I was never able to run Plasma on it, for whatever reason, it always was only partially working, I had a black screen and a cursor, but widgets and panels were missing. But since Arch Linux made Wayland+Plasma 6 a default version I decided to give it a try, and magically it works (xorg still has the same issue).

Actually the article says only part of the story. In fact the hackerhackers has been employed by the independent train repair company, that has a full right to service and cerify trains. The GPS coordinates would deliberately brick the train in their location. The Newag (manufacturer) was loosing public tenders to this independent company, since there a law was introduced that the full train service documentation needes to be available. Only then these "mysterious" breakdowns started to happen. And not the hackers but Newag has serious troubles from side of ABW (counterintelligence and security agency) which is not investigating this story. The story about hackers is more like empty threats from the real culprit.

That's a childish method, and you risk being caught. Much better is to use low statistics experiment real data. If you play enough with data (honest analysis, no cheating), or if you repeat your low statistics experiments often enough, eventually you will find something unusual. Now you can make some nice story out of it. Some reviewers might be picky, but often enough you will get through peer-review. When you controversial paper is published, there will be plenty of theoreticians arguing why you measured something like you did, there will be others trying to repeat your finding. All this plays in your favor - you get citations (h-index) - which is another scientific currency apart from the plain number of papers. Then someone finds that this was just a statistical glimpse, and story ends. Obviously you cannot pull off this scheme too often, but going for it from time to time will boost nicely your records. I know people who build careers on such research.

Assuming you wake up at 7 am, and you have your leisure time starting 6 pm, In a place where I live you have:

• * "Winter" time only:
  • - mornings after/before sunrise 280 / 85
  • - afternoons with/without daylight 169 / 196
• * "Summer" time only
  • - mornings after/before sunrise 218 / 147
  • - afternoons with/without daylight 228 / 137
• * Daylight saving scheme
  • - mornings after/before sunrise 267 / 98
  • - afternoons with/without daylight 195 / 170

Taking into account that I usually get up before 6 am, I would strongly prefer "Winter" time If I had to choose fixed one. But I understand that people have different jobs and habits, so probably equal number of votes would go to the "Summer" time. As a result daylight saving scheme seems a reasonable solution.

Why would you use excel for such data in a first place? There are so many other methods doing that better. Even CSV seems to me like a better option, not to mention something like HDF5.