What Narnia’s Dwarfs Can Teach Us About AI
A stable that wasn’t a stable
Near the end of The Last Battle, the last of the Narnia books, a band of Dwarfs gets thrown through a stable door. They brace for darkness and straw. What’s actually on the other side is Aslan’s country, with sky overhead, grass underfoot and more light than anyone in the story has ever seen.
The Dwarfs don’t see it. They sit shoulder to shoulder in a ring on the grass and insist they’re in a pitch-black stable. Lucy begs them to look up. They call her a liar.
Aslan puts a feast in front of them. They eat it and taste hay and dirty water from a trough. Within a minute they’re fighting over who got the least rotten turnip.
When Lucy asks why, Aslan tells her the Dwarfs picked suspicion over trust. The prison is inside their heads, but they’re locked in it all the same, so afraid of being fooled that nobody can get them out.
Their slogan, repeated through the whole book, is “The Dwarfs are for the Dwarfs.”
I’ve been thinking about that huddle on the grass more than I expected to this year.
The people with forks in their hands
Some people are doing extraordinary work with AI right now, in real jobs, on real deadlines.
I know a consultant who works alone and ships what used to take a six-person team. I know a sales leader whose reps spend their mornings talking to buyers because the spreadsheet work is finished before they log in. Founders are standing up working products in a month that would have needed a year and a funding round in 2023.
I’m in that group, and this is the most fun I’ve had at work in thirty years. My agency does things today I wouldn’t have believed possible three years ago. Our clients get more pipeline from fewer hours of their own time, which is the only score I keep, and that score has moved.
Nobody in this group is a genius. They sat down at the table and picked up a fork.
Five ways to sit in the dark
Then there’s everyone else. When I talk with executives, founders and sales teams, the people who aren’t eating fall into five huddles.
- The ones who won’t try. AI feels risky, or beneath them, or like a fad that will pass. They decided to wait and see. That was three years ago.
- The ones doing the old job a little faster. They use AI to write the same email, build the same deck and run the same process, and they get a 10% gain. Then they announce that the hype was overblown. Nobody in this huddle has asked what they’d do differently if the old constraints were gone.
- The ones whose attention is on the data center buildout. Every conversation ends up at power draw, water and who’s getting rich on the hardware.
- The ones whose attention is on extinction risk. The technology might end civilization, so alarm is the only responsible posture.
- The ones whose attention is on mass unemployment. AI is coming for the jobs, and the sensible response is to brace.
Huddles 1 and 2 are the Dwarfs exactly. The feast is on the table, and they either won’t eat or they chew without tasting.
Huddles 3, 4 and 5 are a different story.
These worries aren’t imaginary
This is where the Narnia story stops matching, and I’d rather say so than pretend. The Dwarfs were wrong about the facts. The stable wasn’t dark. The food wasn’t slop. The people worried about data centers, long-term safety and jobs aren’t wrong about the facts. Data centers consume staggering amounts of electricity. The companies building these models publish their own warnings about risk. Some jobs will change past recognition and some will be gone. Thoughtful people hold these concerns, and parts of what they predict will come true. So I won’t tell you the concerns are fake. What I’ll tell you is less comfortable: you can hold the concern and eat the feast at the same time, and the people at the table have been doing exactly that.
The failure is all-or-nothing
The Dwarfs’ problem was never doubt. Doubt is healthy. Their problem was that doubt became the entire posture. They wouldn’t look, wouldn’t taste, wouldn’t let anyone show them.
That’s the trap waiting in huddles 3, 4 and 5. The concern is fine. The allocation is the problem. If all of your energy goes to the downside and none goes to building, you’ve made the same choice the Dwarfs made. You didn’t check whether the stable was dark. You decided it was.
There’s a practical angle too. The people best placed to shape how this technology gets used, to make it safer and fairer and less wasteful, are the ones who understand it from the inside, and you only get that understanding by using it. A critic who has never built anything with the tools is a Dwarf with a position paper.
Meanwhile the feast improves every month. The tools get stronger and cheaper. The gap between the people eating and the people huddled widens, and it compounds. In a shift this large, the upside doesn’t get handed out evenly. It gets collected by whoever showed up.
Taste it
I’m not asking anyone to drop their concerns. Keep them, and keep pushing the industry on them.
I’m asking for a different ratio. Spend most of your working week building something real with these tools. Take one task you do every week and ask what it would look like with a patient, fairly smart assistant who’s awake at 2 a.m. Then find out. Spend what’s left on the hard questions, informed by what you learned at the table.
The Dwarfs sat on the grass in paradise eating turnips because they had decided in advance what the room would be.
Don’t decide in advance. Look up and take a bite.