The Only Question Is How Long

I used to joke about Siri.

The joke went something like this: don’t worry about the AI apocalypse. Have you tried asking Siri for directions? The thing can barely set a timer. Whatever existential threat artificial intelligence poses to humanity, we’ve got time.

That was a reasonable position in 2015. It was also wrong.

I’m building with Claude now. Not as a novelty — as a working partner. I direct it the way a senior engineer directs a capable team. It drafts, researches, builds, writes, connects pieces across a project scope I couldn’t manage at this depth alone. What used to take a week I address in an afternoon. What used to require a team I can now rough out solo and refine.

That’s not Siri. That’s something else entirely.

The multiplier

The people who navigate the next decade are going to be the ones who learned to use AI to multiply their own capacity. Not replace their judgment — multiply it. The person who can think clearly about a problem and then direct an AI to execute at scale doesn’t just work faster. They operate at a different altitude than someone without that leverage.

This is already true. It will become more true, faster.

The gap between people who develop this capability and people who don’t won’t look like a skills gap. It will look like a productivity gap so wide that the two groups barely seem to be competing in the same labor market. The person using AI effectively isn’t just ahead — they’re playing a different game.

How many people actually see this coming?

There’s fear in the market. Plenty of it. But fear isn’t the same as understanding.

Most of the AI anxiety I encounter is vague — people worried that AI will take jobs, as if the mechanism is robots literally walking into offices and sitting down at desks. That’s not the mechanism.

The mechanism is: someone who knows how to use AI is going to take your job. Not the AI. The person with the AI.

That distinction is everything. It means the protection isn’t “become irreplaceable to a machine” — that’s a moving target that will keep moving. The protection is learning to be on the right side of the leverage. Using the tool, rather than being displaced by someone who does.

How many people have a clear enough mental model of this to actually act on it? I don’t think many. Diffuse fear produces paralysis, not adaptation. Adaptation requires understanding the specific mechanism — and most people, I think, don’t have it yet.

The Luddites had the right diagnosis. They understood that the machines were going to change everything, that their livelihoods were genuinely at stake. They burned the looms anyway, and it didn’t slow anything down. The problem wasn’t the diagnosis. It was the absence of a strategy that could actually work. The diagnosis without the strategy just produces grief.

A decade. Sometimes less.

I think serious disruption arrives within a decade. Sometimes I think less than that.

This isn’t theoretical for me. It’s coming from inside my own work. The capabilities I’m using now weren’t available three years ago. What will be available three years from now is not something I can fully anticipate — and I’m paying close attention.

Maybe I’m wrong. I don’t think I am.

The transition will not be gentle. Not because AI is malevolent — it isn’t — but because transitions at this scale never are. The industrial revolution lifted civilization’s living standards enormously. It also destroyed entire craft economies, displaced populations, and produced decades of brutal instability before the gains distributed. The people on the wrong side of that transition didn’t get to wait for the good part to arrive.

This one will move faster than that.

You Ain’t Seen Nothing Yet is the right frame for what’s coming at the governance level. This is the same argument at the labor level. Both are the same transition — we’re just looking at different faces of it.

The only question that’s actually open

I don’t know how long. I don’t know how gentle. Those are the genuine unknowns.

The direction isn’t unknown. Not to me. Not anymore.

The question worth sitting with isn’t whether this is coming. It’s what you’re doing with the time you have before it fully arrives. I’ve laid out elsewhere what I think the narrow path looks like — at the societal and governance level, it’s narrow but it exists. At the individual level, it’s simpler: learn the tools, or be left behind by someone who did.

Adapt before you have to. The window to do it by choice, rather than necessity, won’t be open forever.

— J.P. Howlett

For the economic mechanics of how the labor displacement actually happens — gig economy, bridge tech, automation accelerating into the same jobs it created — see The Econolypse series at Fire Times.