Early-stage companies are exercises in experimentation, iteration, and figuring things out. You try one thing, it sort of works, but you see a tangential opportunity and go after that. Sometimes that leads to a full-on pivot, other times it leads to an evolution of the opportunity.
This process can create multiple products, services, projects, and it can look messy. I love it when founders bring it all back together into one cohesive package. That doesn’t always happen, but when it does, it is a thing of beauty.
That happened this week with our portfolio company Numerai. I saw this tweet by Numerai founder Richard Craib:
If you have not been following Numerai, let me explain.
The initial idea behind Numerai was to build a quant hedge fund using “the crowd” to surface the machine learning models. They do that by running the Numerai Tournament and taking the best models and operating a hedge fund with them. That is what Numerai does and that business is working well.
Along the way, Numerai observed that incenting the players in the tournament with a crypto-asset they could stake on their models was a better way to crowdsource the best models. So they made a crypto-asset called Numeraire.
Numeraire is a crypto-asset that you can trade, own, and, if you are playing in the tournament, stake.
So then Numerai was in several businesses, the hedge fund business and the crypto business.
And then the team observed that there was a generalization of the thing they had built that had much broader applications – a decentralized marketplace for predictions. So they created a protocol called Erasure that allows anyone to build markets for predictions.
It started becoming a fairly messy story.
But last week it all started coming back together when team successfully ported the Numerai Tournament to operate on the Erasure protocol.
This accomplishes several things. First, it starts to tie all these loose threads together and the story becomes easier to understand. And also, you have a very successful application, the Numerai Tournament, running on Erasure. So developers can see what Erasure is good for and can start building other things on it.
I am excited for the Numerai team. And I quite like this example of how things get messy as you expand the opportunity but you can clean things up by bringing it all back together and would encourage founders and teams to look for opportunities to do that whenever possible.