I suspect that, in practice, what many enthusiastic advocates mean by “learning AI” is actually “learning to need AI”.

In other words, the aim is to get kids used to using AI as soon as possible, so that they do not learn the skills to function without depending on it.

You absolutely need prompting skills to use AI usefully. You need to know how to eliminate sycophancy, how to ask for and check primary sources, and how to use follow-up questions.

I've been using AI for some legal issues, and it's been incredibly good at searching for case law and summarising the key implications of various statutes - much more efficient than web search, with direct links to the primary sources it finds.

I'm still the one gaming out "What if...?" and "Does that mean..?" scenarios and making sure the answers are grounded in the relevant statutes, and aren't mistakes or hallucinations.

It's not so much a prompting problem as a critical thinking and verbal reasoning problem.

As someone who sells AI... You'd be shocked at how bad people are at using AI.

My 6 year old kid who watches me is a better prompter.

Especially since kids these days aren't even very good at using computers:

http://www.coding2learn.org/blog/2013/07/29/kids-cant-use-co...

It seems to me that if someone can read and think critically-- they can RTFM and get much better much quicker at computers and AI than people who spent all their time tapping an ipad to watch the next video.

Presumably prompting skills like https://github.com/califio/publications/blob/main/MADBugs/CV...

It's true that you can use LLMs as a learning resource and to unblock you. But students just aren't. They are using them as a way to avoid thinking, avoid research, and just spit out an answer they can paste in to their homework.

I assume "AI workflows" means knowing how to split up a task to create a chain of agents that can complete a specific task reliably.

A bit like software development.

Yeah I'd be happier if they learned how an Apollo computer worked (even though it has virtually no relevance) than how to use Excel.

Just as "there is no royal road to mathematics", no AI can do your learning for you. The need for memorization of essential math identities (like multiplication tables and use of fractions) or rules of grammar (like verb conjugation or use of anaphora) will never be enhanced by AI. There is an essential role for good old fashioned rote learning that can't be avoided. To pretend AI will not impede that learning is a fool's errand, literally.

If AI is still too stupid to show people how to work with it, and to notice their lacks and anticipate their needs, it can't have become that indispensably useful.

The entire point of AI is to accommodate the user. AI doesn't do anything that people can't do, is worse at most of those things, but is a lot faster at some of them (basically looking up things.) The point of AI is natural language UI.

Teaching people how to use AI is just teaching people enough about the world to give them something to ask AI for.

And s/he was right. Most students who were brought up with calculators in math class cannot do basic math without one today. When shopping in groceries, they have no idea if one product costs more than another by weight. They're easy to bamboozle with the simplest misrepresentations of numbers. Is one choice of product really better than another, fractionally, or corrected for a shifted baseline? They don't know and can't use basic algebra to find out.

This is bad -- an F grade for the education system that let them slide by without learning an essential skill. The chinese aren't this lazy. And if we persist in not learning this, America's future will regress to us asking them, "Do you want fries with that?"

AI is important but we don't know what skills will be relevant in 10+ years to harness AI (I can't imagine prompt engineering is much the same). Anyway, would a typical teacher be ahead of the curve on what pedagogical tack to take here even if it was appropriate?

The best thing to do is to set the kids up to learn the most important thing - which is how to teach oneself. If a kid can read about something, and then understand what was important from the reading, and then write about it, and then know where to go next they will be well served in the AI world.

AI is here to stay. But learning to copy-paste homework into a chatbot is not really a skill one needs to learn.

You can put a candy bowl in front of kids and tell them not to touch it. Or you can just not put it there. Ultimately kids will be less distracted when you remove the source of distractions. Phone bans in schools are showing this already.

blaming discipline is how we got here. these devices are engineered by teams of psychologists to maximize engagement. expecting a 12 year old to resist what grown adults with PhDs can't is just setting kids up to fail. removing the distraction source is the rational move.

Most university students still behave like kids. I don’t think you can expect under-20 students to behave as adults, honestly. I went to university again around 30 . And my wife teaches first year students. Maybe I am just old now but those students are just kids and behave like large children.

Transitioning to adulthood but there is still a long way to go for most.

This is written with what feels like the peak understanding of my kid's school's IT department: "well, they're just so smart, we can't find any way of stopping them!"

An allowlist might be a good place to start.

Yes, but you aren't going to give your teacher a remarkable to take home. Hence a sheet of paper comes in handy :)