People using AI as a crutch will never do anything other than get themselves into trouble eventually if they don't know how to spot check results. If something AI responds with seems off, there's a good chance it is. Want to use AI to do menial tasks like "create an array where each entry is the name of regular file found in a specified directory" because you don't want to remember the language syntax of every scripting language available is fine. It also does an excellent job of taking large documents and providing summaries. But asking it to do "research" for you? You'll end up getting in hot water because it could generate false results. Always verify.

Yeah, I've seen the various videos and instructions, but the trouble is that it makes certain other applications fail such as the Microsoft Store, the search bar in the start menu... so if you can live without those, you'll be fine. On the plus side, it removes Bing searches.

... and do so while leaving the remainder of Windows actually usable instead of crippled... will be a hero to a lot of people.

... that overpromises and underdelivers... what a surprise.

Make the stuff they write start to interconnect as the semester progresses too. It will help them learn a lot more about how to make reusable code, common libraries, etc.

The only nit to pick here is that while gatekeepers might reduce misinformation spread, it is more about who's paying the gatekeepers and how misinformation is defined. It's more accurate to say that gatekeepers drastically reduce misinformation spread based on how misinformation is defined for them by their bosses, whether those bosses be a government entity, an interested corporation, or some other special interest group.

Maybe Iran could start up their own AI farm and create an appropriate LLM that understands that custom. Of course, that would limit the overall knowledge base and usefulness of it for anything outside of cultural interactions...

... everyone who has an account there went in and disallowed use of their accounts for this. But as others have noted, it's Microsoft... so who knows if they will actually honor it or not.

AI is mostly a toy, but it can also be a useful tool. Unfortunately it's becoming a crutch and something that's taking away critical thinking and research skills. Remember this one from about 6 weeks ago? Anything AI gives you as a result must be verified.

Eating too much sugar is just as bad and can cause problems as well. Try looking at the nutrition labels on packages... compare a "fat free" version of a product to its regular counterpart. Fats bring a lot of flavor to food... so making it "fat free" means they make it bland. The only way to bring flavor back is to add sugar, so that nutrition and ingredients label will reflect that sugar was added. Sugar is addictive too. I refuse to drink "diet" or "no sugar" drinks. I'd rather get 100% fruit juice and dilute it with some sparkling water.

Always!

... hire a few young people for these soon-to-be retirees to mentor and pass on some of this institutional knowledge?

Going through German as a refresher, what I find very wrong about the way they teach it is when it comes to nouns... you would never learn it as Hund, Katze, or Kind (dog, cat, child)... you would learn der Hund, die Katze, or das Kind (the dog, the cat, the child) because you need to learn whether it's masculine, feminine, or neuter, otherwise you'll pick the wrong definite or indefinite articles in different situations.

Beyond a local neighborhood-maintained book or some kind of advertisement from my county of residence in a "Best of X County" type thing, I haven't seen an actual newly updated phone directory in a couple of decades. I remember delivering them as part of a Boy Scouts fundraiser back in the day too. White pages, yellow pages, and "Let your fingers do the walking"... it wasn't difficult... but then again, we also used to go to that big wall of tiny drawers to search for a book, author, or subject. :)

We have AI making up court rulings that sound completely legitimate but a simple Westlaw or Lexis-Nexis search would tell you that there was no such case... AI will gladly spew drivel about one side of a political argument, but when asked to spew similar drivel about the other side it will say, "I'm sorry but I cannot delve into politics"... and now they want ads in there? In baseball that would be your three strikes and you're out.

AI is a toy. Treat it like one.

This is fine and dandy... but who controls the accreditation agency? We have seen plenty of times in the past where junk science was pushed as part of the dictatorial takeover of Russia, Germany, China, and North Korea. At first, it's to help swing the populace to the side of the rising dictator... but once in power, it's about "the official story" to keep them there. Go against that official story and you could end up in a lot more than hot water.

What needs to be fixed is the lackadaisical peer review process. Once upon a time, you had to rigorously document your methodologies... how you collected and analyzed the data, and how you came to your conclusions. These days, too many peer reviews are glossed over and an insufficient number of attempts are made to repeat any part of the process, actually go through the data and analysis process to see if those conclusions are correct, etc. It leads into some of the stuff others posting here have said: "publish or perish" mantra.