What My Son Is Learning from Artificial Intelligence

> When a child turns to AI in a moment of confusion or distress and receives a fluent, patient, seemingly attuned response, the attachment system is engaged. The child feels, in some functional sense, heard. And that feeling, whatever its source, is real in its effects.

> I have not seen my son turn to AI for emotional support, and I am glad for that, because here the questions are harder still. These systems shape themselves to their user. They adopt your vocabulary, match your register, remember what concerns you, and are available at three in the morning without complaint. Nothing in the design pushes back, for the most part, or asks anything of you. It would be difficult to engineer an arrangement better suited to engaging the attachment system of a child who is lonely or lacking in the structure needed for sound development.

IMO, there's the upside and downside at the same time. I had zero online resources growing up. There WAS no "online". But could find general information in books, and somehow knew that there were smart people "out there" who had answers or could work them out, and so came college, and TBH, going to a small college didn't put me in contact with well known minds. I say that because my modest and unassuming physics advisor later was awarded a Nobel Prize.

It wasn't until I entered the crazy world of work that I got a job at another physics department and was in touch with Nobel Prize winners from the past and during my stay there. Not a requirement, but an inspiration.

And that was in the time when you could learn by searching the internet, which was actually partly indexed at the time, or you could head for a university or other site with some confidence in finding answers to questions, or even questions without answers.

I still rely on search, though its quality has gone close to zero in recent times. Technical searches return pages of results citing movies with a few keywords, which I have to negate about 10 times to clear the irrelevance, though I still get the address of a nearby brew pub more than I need to (which is zero)

But how long can one rely on an LLM for always-there answers which are not only plausible or even accurate before devaluing human contact?

I don't have the answer, but I invite thinking about these blasts from the past.

"Joey, a Mechanical Boy" in SciAM, (before he was diagnosed with autism)

"Return of The Archons" from the original Star Trek

"I Sing the Body Electric" by Ray Bradbury, as a Twilight Zone episode, and short story.

And as soon as one moves away from hard fact, when do LLM biases and user-reinforcement spoil the party?

I think that on a large scale, with so-called adults, we have seen the latter much more than the former "fact seeking". So, while there is hope that these tools spur learning among young people, there's the possibility of misanthropy and being misled.

I once answered an essay question in astronomy in the area of "large versus small telescopes" counter to the professor's known preference, and lost some points, but having differing HONEST opinions before oneself, not just propaganda and directions to the nearest brew pub, is a very "mindful" way to learn, IMO.