The Tiger Equilibrium
I don’t think I’ve ever told my younger siblings off for not working hard enough. I’ve made a wry comment about an incorrect Latin conjugation; I’ve made desultory attempts at tutoring them; I’ve encouraged them to get a summer job ; I’ve bought them books, subscribed them to email lists, and even, in one case, bought them a laptop to encourage them to learn to code. I’ve set an example, perhaps, but never offered overt criticism or direct feedback. Certainly, if I did, I felt bad about it afterwards.
Each of them have been successful; each, trivially, could have been more successful. My sister never got that summer job. I don’t know if my brother read those books, but I know he doesn’t read those emails, and my other brother certainly didn’t learn to code. We never kept up the tutoring sessions. I feel every one of these as my failure - perhaps a shared failure, but certainly a failure. I’m sorry that my brother didn’t learn how to code. I’m sorry that I didn’t try harder to encourage him; as I write this, I feel a guilty urge to try again.
The other day, I watched a friend shout at his brother to work harder, not be so lazy, get out of the house, produce better work.
I was shocked; it wasn’t that I disagreed with the sentiment - the previous paragraphs should demonstrate that - but I couldn’t imagine communicating that message, in that tone, to my own siblings. In my family, parents make their children do their homework; but they don’t force them to work hard, to produce exceptional results. Is that bad?
This attitude is driven by fear of backlash. If you force children to do anything, but especially work, they will respond with mulishly stubborn refusal, a determination to do exactly the opposite of their parents wishes. Rather, children should be left to become who they are, fulfilling their destiny as the Disney movies would have it. Under compulsion, children become Bart Simpson or Bartleby the Scrivener; perhaps my parents have faith in the Ciceronian maxim that “what best becomes a man is whatever is most peculiarly his own”. There’s some truth to this outlook. I was certainly the sort of child to do exactly the opposite of what he was told - but it seems like that should be the exception rather than the rule. As a pattern of behaviour, it’s hardly Confucian.
There’s a spectrum of parental compulsion: Padraig Harrington at one end, Tiger Woods at the other, along with John Stuart Mill, Whiplash, and the Tiger Mother. But if parental compulsion is a spectrum, it’s an also equilibrium quantity. As you move up and down that spectrum, you’re making tradeoffs: you increase the risk of burnout: the child turning to drugs, running away from home, falling out of love with the viola. Perhaps strictness makes children less original - this essay about Singaporean education bemoans something like that; perhaps children can only work hard if they’re free to discover their passions.
Mobile phones, “the algorithm”, and AI all shift that equibrilium. The distractions are more potent than they used to be. But just as importantly, the shortcuts are better too. It’s never been easier to not do the work.
I feel this viscerally. All the skill I was built on sucking at stuff that didn’t work - writing bad essays, losing competitions, rote learning, debugging code - and every time I sucked, I was choosing to suck over doing something else - something easier, perhaps something more alluring. People ask me how I learned to write, and I think back to being sixteen in a medieval history class, writing a succession of bad essays. I couldn’t get the structure right; I didn’t know how to order my paragraphs, how to write a topic sentence, how to leaven my argument with evidence. I kept turning in these misshapen screeds, and over time, I got better.
To gain that skill, I had to traverse a J-curve; I had to choose Type 2 fun over Type 1; but Type 1 fun is getting more fun everyday. It’s not longer a Marshmallow Experiment; it’s an Ice Cream Sundae experiment, except that the sundae isn’t sitting on the desk beside them - it’s being trolleyed through the door, in unlimited quantities, by hundreds of billions of dollars of compute. If Claude is better at writing Python than I ever will be, then why bother writing bad Python? Why not just let the AI take the reins?
Short of AGI, the impact of AI that I’m most worried about is developmental. How do we create space for children to learn, when AI can provide answers on demand? How do we ensure that children can be bad at things, so that they can get better?
Niall Ferguson’s concept of the Cloister and the Starship is a great answer, and one rooted in the traditions of Oxford and Cambridge. Handwritten essays and tests, in-person tutorials and competitions - this will become more important than ever. But enforcing the separation between Cloister and Starship will require the equilibrium of parental coercion to shift. Parents who want the best for their children will have to become more demanding - more Tiger. Adults who wish to continue learning - continue to outpace Claude - must also become more demanding of themselves.