Boring is interesting
A couple of weeks ago I published my 'Boring AI' Manifesto.
I had a bit of fun with it, but I believe it's important. It's a core idea in my upcoming book, The AI Pilot Handbook.
A few people asked why I'd use the word "boring" at all.
Boring is interesting
Boring is interesting, because it has a couple of meanings:
The first kind of boring is the one we all dread.
It's tedious. It's laborious. It's copying data back and forth between two platforms. Filling in a long online form the same way every time.
Copy. Paste. Save. Repeat.
It's repetitive, it interrupts and steals time from valuable work, and it eats away at your motivation and energy.
The other kind of boring is what we want from systems we rely on.
I want my seatbelt to be boring. I want my trans-Atlantic flight to be boring. My smoke alarm to be boring. The same for my accounting software (although I'd rather never have to interact with that at all).
Boring here means predictable, consistent, safe.
No surprises.
Surprises in these systems are usually expensive, sometimes even dangerous.
This is why I think boring is underrated. I want to build AI that's boring in the second sense so
Remove the tedium with reliability.
If the AI does its job so well you forget it exists, that is not a bug. That's the goal.
If you want to try this without making it a big AI programme, start small. Pick one tedious task that happens every week, automate it, and measure the hours you get back. Then do the next one.
If you'd like to know more about The AI Pilot Handbook, let me know in the comments and tell me what you're curious about.