Positive sum opportunities for AI
I believe most people assessing AI opportunities look in the wrong place.
We live in a positive-sum world (thankfully). It's exactly those opportunities that we should be looking for with AI. It's what will make humanity prosper.
What do they look like? How can we find them?
Where the profit flips
The key question to ask ourselves is
what was unprofitable with traditional labor but is now profitable using AI?
Let's take a look at an example:
| Traditional labor | With AI | |
|---|---|---|
| Revenue opportunity | $10.000 | $10.000 |
| Cost to capitalize | $12.000 (salary) | $6.000 (tokens) |
| Projected profit | -$2.000 | $4.000 |
Kablam! That's newly available $4.000 in profit with no job loss!
But there's another kablam hidden here. Hiring new human labor is time-sensitive. Consuming more tokens is near-instant!
So once you find such gigs and they're reproducible, man you are ON FIRE.
Just sign a 1000 more gigs; consume 1000x tokens and make 1000x the profit.
KABLAM that's $4 millies for you!
My gut tells me there is so much of this opportunity is still laying around.
We humans are bad at the zero-to-one. We try to make existing profit margins increase first. We instead should be creating profit margins that didn't exist yet!
There's more!
As costs of AI go to zero while the reliability of AI grows, this opportunity grows further.
More opportunity that wasn't unlocked before!
We will get better harnesses and better models. That means AI can finish more complex tasks with satisfying results.
Which means we will have more profitable opportunities that were unprofitable before.
More pie!
The banger of it all
We are restricted by labor. It's necessary, but expensive and hard to gather.
You find a profitable gig, you want to do it 100x, but you can't hire fast enough. Labor shortage is the word of the decade.
With these new opportunities, that constraint evaporates. Scaling means increasing your API spend.
So you get two compounding forces at once:
- The opportunity space is expanding: as AI gets cheaper and more capable, more previously-unprofitable work crosses the profit line.
- The scaling constraint is gone: once you find a winning formula, you can 1000x it tomorrow morning.
New markets opening up and near-zero friction to capture them.
So let's do it! Let's 10x our economic output. Go digging, build these next zero-to-one companies! Grow our pie!