University halls, 2am. You’ve been there since 10. The essay is due at 9am. You keep telling yourself the next 15 minutes will be when you finally start.

When you force yourself to face the word document, there’s a feeling of physical excruciation. Not boredom. Not laziness. Something more visceral. Anyone who’s been there knows exactly what I mean. Anyone who hasn’t probably thinks I’m describing a character flaw.

I don’t have a formal ADHD diagnosis. You could argue I’ve exhibited the symptoms through school and university. Whether ADHD is a first-class neurological citizen or a societal hallucination about which personality types fit modern expectations, the experience was real.

Here’s the strange thing: since LLMs, it’s gone.

The Nintendo effect

Work has become so enjoyable it’s almost indistinguishable from playing Nintendo at age 10. Identifying problems, solving them, improving solutions, making things robust. At a rate that’s almost perfectly tuned to a dopaminergic feedback loop.

Steve Yegge describes vibe coding as addictive. “Random reinforcement for the win.” He’s talked about having to drag engineers off stage at conferences because they can’t stop. He codes during meetings now. The slot-machine pull of every query, the dopamine hit of each successful change.

I recognise all of this.

The impact compounds over time. Because it’s so easy to improve things, the perceived opportunity cost of stopping is high. But it doesn’t feel unpleasant. It feels like there are more coins to collect, more mushrooms to grab, and the rate is compounding.

The burnout question

I’d always assumed burnout comes from working too hard. That cringing, sapping inability to face any problem because you burned the candle at both ends.

I haven’t found this to be the case. Periods of underemployment have been harmful. The current state of being functionally overemployed, diving into work because it’s genuinely fun, hasn’t produced the expected symptoms.

I mentioned this to someone in a management role once. The urge to work on weekends because the work is dopaminergic. They couldn’t conceptualise it. For them, working more meant increasing burnout risk. Full stop.

That connection made sense for the old kind of work. I wonder if the norms that established it are being undermined.

Like attracts like

A quirk of social life: the people I talk to most over WhatsApp all exhibit the same thing. The same relationship with AI tools. The same acceleration.

The ratio of people I know who have interesting things to say about this stuff, who also have an Adderall prescription or who are dead ringers for the thing I’m describing, is disturbingly high.

These aren’t early adopters in the usual sense. It feels like certain aspects of how we think became failure modes in modern society. Sitting at 2am unable to start a thousand-word essay on Descartes. For reasons not yet clear, LLMs let you skirt around the problem.

The idea that you can flick on dictation, or Super Whisper, and the problem of starting is just gone. The problem of distilling and refining becomes frictionless. You’re still in the loop, but everything is greased.

The personality question

I’ve written before about how it takes potentially a decade or two to catch up with a big technological change. There’s an overhang in terms of form factor and metaphor.

I suspect there are implications for personality types that will have decades of overhang too.

If you froze the current tools long enough for people to actually adapt, would the typical software engineer personality still be dominant? There’s a family resemblance among software engineers, a distribution that skews a certain way. Would the same personality type thrive?

It feels like a very open question. The fact that it’s unexplored makes me think we’re in for a surprise.

I’m not claiming the world is getting better. No free lunch. But for a minority, things have quietly turned upside down in ways that probably aren’t obvious to anyone else. The student who’s never stuck with a blank word document at 2am.

I wonder if this is the only such group. And if not, whether the implications will fan out from here. This is bigger than a butterfly flapping its wings.