Live Stream: How to Start a Hard Tech Startup with Sam Altman, 6:00 EST

Why is this being streamed from an unknown YouTube account vs. an official YC/MIT organization?

EDIT: Stream is actually real

Preliminary notes jotted down on the fly. If something seems odd or silly, I probably didn't hear well or mixed things up.

YC: - 11 years.

- 1200 startups.

- Recently started to fund "hard tech" (few are funding them).

How to start a hard tech company:

- What do you need to do differently than a software startup:

    - The least amount possible. You're not that different.

    - Same principles apply.

    - They were first segregated at YC. Turned out to be bad.

    - Most important change:

        - YC & the market don't care: Create value or fail.

        - Don't think you're that different:

            - "But my idea takes 10 years!":

                - What can be done in 3 months?

                - Consistent incremental wins compound. 

                    - 10% every week for years.

                    - Find something small and get it done.

- Advantage:

    - 11,000th photosharing app is hard.

    - Competitive advantage (4 nuclear fusion companies in North America).

        - Tesla, SpaceX, etc. cf. Michael E. Porter.

- Pay attention to what software companies did well/right:

    - If in doubt about a field, 2 important things (good indicators):

        - Cost down.

        - Cycle time is getting shorter.

- Difference:

  - SW companies are usually "clearer" than HW.

- It's a great time to start a HW company. Make sure:

    - Weekly wins.

    - Keep costs down.

- Make peace with the "world":

    - "Big companies don't play by the rules":

         - Make peace with that.

         - No excuses.

         - Your duty as a founder to get that 10%/small wins consistently.

         - Keeping cycle times down is important because:

             - Maintains *momentum*.

             - Avoids mistakes (typically made by bigger companies).

         - Establish a culture of winning from the beginning, no matter how small.

         - Momentum attracts good people:

             - People want to join a winning team.

         - How?:

             - SW startups drive cycle time down by getting closer to the customer.

- Detours:

    - Sometimes you can't reach "final" goal directly.

        - ex: Tesla - Roadster (revenue) --> Model S --> Model X --> Model 3.

        - Not from 0 to mass produced car. SpaceX: Not from 0 to colony on Mars.

    - Even if you're not in YC, you can adopt this model.

    - More people fail because of "grandeur" than because they did small things.

- Raising capital:

    - Too much money can kill you. (ruins your focus, for example).

    - Statistically (YC):

        - Companies that raised $40MM tended to fail.

- Business side:

    - Don't be all hardcore science and bad at everything else.

    - Someone should take care of the business side.

    - Not hard: you can do science, you can do business.

    - It can be boring for some, but it needs to be tended to.

- Many will do the thing everyone is doing. Think of a worthy problem. (?)

- U.S. Context:

    - Centuries of abundance and massive growth.

    - Not so much recently (2%) --> Democracy works badly in this context.

    - The only hope is innovation (and a more important growth).