“I don’t have enough experience to become a founder.”
“I’m not ready to become a founder.”
These are some of the common reasons we hear why someone can’t be a founder right now. Many potential founders don’t think they have enough experience, or aren’t ‘ready’ to be a founder.
No-one is ready to be the founder of a billion dollar company on the first day after graduation. But the same is true of being the CEO of Google. In both situations you need to gain experience to achieve your goal. The important question is where do you gain the experience that will most improve your chances of success?
What experience would you give a computer if you wanted them to found a company?
Let’s run a thought experiment – if you were trying to get a computer to build a world class startup, what would you do?
It would make sense to use Reinforcement Learning (RL) as there isn’t a data set for you to train your agent on (as a founder is creating something that hasn’t existed before).
To help the agent achieve its goal by learning through trial and error, you need to carefully choose the environment and have a reward function that generates the right behaviour.
You would give the agent an environment that allowed it to iterate based on experimentation with customers. The goal would be to build a product that people engaged with (not sufficient to succeed as a founder, but a necessary starting point). Over time the model would interact with customers and gain experience that would allow it to improve its product over time. Even if the model didn’t succeed initially, it would be learning the skills and behaviours to improve its capabilities over time.
For a computer getting the environment and reward function right is critical to its learning. The agent will learn faster when the environment is most similar to what founding in the real world is like.
Would you give the agent an environment that mimicked a big company? This sounds like the right environment for the future CEO of Google; where they can learn how to navigate existing organisational structures, technology and politics. But, it’s unlikely that this environment would enable the agent to learn the required behaviours and skills to succeed as a founder, particularly when the reward function of a corporate employee is to get promoted.
This seems logical when referring to a machine and yet many potential founders still believe that getting non-founding experience (in a corporate or startup) is essential before starting a company.
You will never feel ready, but that’s ok
Working at a big company, whether it be Meta, Alphabet, Bain or Goldman, is a great choice if it aligns with your career goals. If you are a highly ambitious young person, who one day wants to be the CEO of a multi-trillion dollar company, it’s worth starting out in a corporate environment. You will gain relevant experience that will hone your future CEO intuition.
However, if you want to be a founder you need to gain relevant experience from the relevant environment with the right goal. This is the key principle behind the Founding Career. You need to be able to iterate and experiment within an environment that allows you to develop as a founder. As with the RL agent, each setback and success becomes a data point that feeds into your specific knowledge about the product you’re creating, and your general knowledge of how to be a founder.
You won’t be ready to be Patrick Collison today, the founder/CEO of multi-billion dollar Stripe, but you are ready to be the Patrick Collison of 15 years ago when he was founding his first startup, age 19, that exited for $5m.
As an ambitious young person you want to and should be on an insane learning curve, just make sure you’re learning from the right environment.