
When I look to an incumbent, I ask what camp they fall into. For our own business at @tryunusual, I ask how we stay fluid enough to consistently remold as the world changes around us at a blistering pace.

happy sisyphus, cofounder @tryunusual

When I look to an incumbent, I ask what camp they fall into. For our own business at @tryunusual, I ask how we stay fluid enough to consistently remold as the world changes around us at a blistering pace.

And cynically, 3) Those accepting their death and dressing up their income statement to look good while the bombs fall.

Second, for existing companies, we can put them in three groups: 1) Those letting AI spend grow out of control without intent to reshape are likely doomed to fail. 2) Those swallowing the bitter pill that they can’t spend their way out, and that they must remold around AI.

First, if you have any doubts of the productivity future AI will yield — look to these high growth companies born in the last two years. Many of the highest growth startups are spending as much on tokens as payroll.

What does it mean pragmatically today? Everyone is worried that large companies cutting AI spend indicates the technology won’t deliver on its promise.

Businesses that retroactively adopt a technology must make the bold decision to totally *reshape into* it — expensive, distracting, and often carrying repugnant conclusions.

This is history's rhyme with today’s AI rollout. Businesses born alongside a technology grow *around* it — for the cotton mills, the winners were "jenny-shaped", today the winners are “Claude-shaped”.

The boost is explained in part by the literally shape of the newer businesses. Marginally wider building plans could accommodate two mules back to back, such that a single worker could operate both simultaneously. The winners were "mule-jenny shaped"

In 1806, factories under three years old produced 60% - 80% more per worker than older mills that were *also* mechanized, controlling for machinery type and size of the factory.

My favorite way to think of froth in the AI market is through the lens of 19th century French cotton mills. This period saw the introduction of the mule jenny, which massively multiplied a single worker’s ability to produce thread. ideas.repec.org/a/ucp/jpolec…

When I look to an incumbent, I ask what camp they fall into. For our business @tryunusual , I ask how we stay liquid enough to consistently remold as the world changes around us at a blistering pace.

And cynically, 3. Those accepting their death and dressing up their income statement to look good while the bombs fall.

1. Those letting AI spend grow out of control without intent to reshape are likely doomed to fail. 2. Those swallowing the bitter pill that they can’t spend their way out of a structural problem, and that they need to deeply remold themselves.

First, if you have any doubts of the productivity future AI will yield — look to high growth companies born in the last two years. For existing companies, watch to see what of these three groups they fall into:

What does it mean pragmatically? Everyone is worried that large companies cutting AI spend indicates the technology won’t deliver on its promise. At the same time, we see many of the highest growth startups spending as much on tokens as payroll.

Businesses born alongside a technology grow *around* it — for the jenny it was physical, today it is businesses being “Claude-shaped”. In contrast, businesses that adopt a technology must make the bold decision to reshape *around* it -- often carrying a repugnant conclusion.