Courage is the numerator, and the denominator is exploding
What happens when "brilliant thinking" is no longer rare?
Peter Thiel famously wrote, “Brilliant thinking is rare, but courage is in even shorter supply than genius.”
What happens when brilliant thinking is no longer rare? Imagine every person granted artificial intelligence that radically improves their agency, IQ, and speed of execution…
I think intelligence loses competitive value, and the premium on courage rises.
In Silicon Valley the last three years have seen developer productivity more than double, and organization-level productivity increase even more. Hardware is seeing similar, if lagging gains.
The time required for a founder to become an expert in a new industry or discipline has collapsed. VCs who might have relied on a “we don’t understand that space” justification no longer have as strong an argument. The truth is, you can just do things in a way that was impossible before.
At Enhanced Radar, we recently engaged a highly-recommended RF engineer who was involved enough in the invention of Wi-Fi that he now flies his own aircraft with his age-gap wife. He proposed a solution for a specific technical capability we needed, but then gave up after a month of work, telling us it was impractical. Over a week, our CTO used AI to digest decades of hyper-specific tribal knowledge locked in PDFs and forums to develop firmware that a domain expert couldn’t do in a month.
In other words, new AI capabilities change not just the velocity of building, but can functionally flip the binary from impossible to possible, even when the final product involves no “chat interfaces” or “agents.”
Strangely, this newfound leverage has not resulted in a commensurate increase in courage. Ambition, maybe. Good, but table stakes. The courage Thiel talks about is the courage to seem wrong, or to stake a bet on the future more firmly than those who pick their vision of the future from some prix fixe menu of narratives currently sloshing around the discourse. In Thiel’s words, “it’s psychologically difficult because anyone trying to answer must say something she knows to be unpopular.”
It hasn’t yet been canonized, but what may go down as the most consequential act of the current AI cycle had nothing to do with model weights or arXiv manuscripts — it was OpenAI’s decision to seek $1 billion from Microsoft in 2019.1 While many beliefs and personal traits were necessary to make this happen, the critical one seems to be balls.2
In a world of multiplying intelligence, even execution will be commodified to some extent.
Courage, relative to intelligence, is in ever-lower supply. And the reward for it has never been greater.
To oversimplify and to retcon somewhat, you could say this deal was borne out of the conviction that scaling laws would hold. And in 2019 the Hacker News crowd was not buying it. And it won’t
Loosely speaking, Google (and most of the other hyperscalers) had all other ingredients besides balls.