What skills should kids build that AI can't replace?

Ryan Kiick does not treat this as a yes or no question. For writing that is an assignment rather than a passion, on a topic the student has no interest in, he says it is okay to have a little bit more bolstering of tools along the way, and he compares it to using a calculator to take off some of the heavy lift. The line he draws is mastery. Where the point is to actually acquire the skill, letting AI stand in for it is the problem, because that is not offloading something the student owns.

Kiick's list is weighted toward anything real time. Strong communication, real-time agility when you do not have immediate access to AI, coming up with an idea and overcoming objections to it, resiliency, grit, and executive functioning. He expects the value of one-to-one, real-time human connection to hold for quite some time. He is also willing to say which way the other side goes: the ability to do something like an algebraic proof may diminish, because that can be offloaded.

The pattern Kiick recommends is to use AI for its questions rather than its answers. You state the problem and ask the model to interview you, which keeps the reasoning with the person. He pairs that with giving students tasks where the output can only come from them, since AI cannot write about an experience it does not have. In his words, AI helped his team but it did not replace them, it made them better in that moment.

Kiick says a lot of things could happen and that it could go the way of the dodo, but he does not predict that it will. He points out that every applicant could always theoretically have had someone else write the essay, so this is not new, only easier and more accessible. His guess at the first wave is colleges pushing the prompts to be even more highly personalized. After that he wonders about a recorded answer instead: you log into the school's platform, you are given a question, and you get thirty minutes to tell them your answer, ums and all. As he puts it, he is pretty sure you are not getting help on that one.

Kiick says there are studies showing that physically writing your own notes commits things to memory far more effectively than sitting in a room taking notes on a laptop, and he treats an automated notetaker as a further step away again. He does not cite a figure. The distinction he draws is between passing the test and still having the material: review an AI summary and you can do well on the exam, but come back to the topic a month or two or three later and it is probably gone.

Kiick defines it by its parts: prioritization, organization, motivation, and the ability to move things into working memory quickly and efficiently. It matters because it is the layer AI does not supply for you. He also makes a business observation about it. Everyone seems to talk about executive functioning, a lot of parents think their kid probably needs it, and most of them still do not go seek out an actual intervention.

Kiick's position is that the outcome is tied more to the major than to the individual institution, and his blunt version is that out of the gates an accounting degree from a state school beats a sociology degree from Harvard for getting a job. Where brand does carry weight is at the graduate level for certain careers. He says that correlation has never really been proven at the undergraduate level. He also likes to point out who actually has the lowest admit rate in the country: you, because you can only attend one college out of thousands.

The test Kiick offers is time. Come back to the same topic weeks or months later, without having used it in between, and see whether it is still there. If it is gone, what happened was short-term transactional learning. Nick makes the related point in the episode: if school is oriented around absorbing just enough to do well on a test, the incentives are not pointed at long-term retention, and he thinks the structure needs adjusting so they are.