1. Why not to start an AI startup:
    1. lots of competition makes differentiation / sales hard
    2. young, unstable field which changes rapidly
    3. ML can be tricky
    4. OpenAI etc have massive headstarts
    5. uncertainty about future capabilities (x-risk, obsoletion, AI winter)
  2. Why to start an AI startup:
    1. AI summer about to start / started
    2. even if capabilities plateau, current models can do so much
    3. this is a major movement, like the introduction of the internet
    4. lots of money and interest around
    5. potential for greatness, robot king
    6. very philosophical, not boring like brex or resend
  3. Why I personally should not start an AI startup:
    1. not super sciencey
    2. never ran a successful company before (do something easier!)
    3. sensitive, lots of negative feelings ahead
    4. hate AI hype and charlatans
    5. young, somewhat inexperienced
    6. would obliterate my free time / freedom, no musical theatre writing
  4. Why I personally should start an AI startup:
    1. (all of point 3, lol)
    2. very very enthusiastic about ML
    3. somewhat wide breadth of tech knowledge / moderate ML skill
    4. want to put youth, time, energy into something meaningful (and make money)
    5. possibility of leading to other interesting things like robotics, politics
  5. What kind of AI startup would I not start?
    1. compute is a no-go, hardware is not startuppy and software optimisation is too hard
    2. fad company (like dreambooth avatars) is fun but not long-lasting
    3. B2C plays are not easy
    4. not interested in something that only peripherally relates to AI
    5. probably not API over open source as that races to bottom on pricing
    6. don't wanna be tied into one field like law or drama or automotive
    7. would prefer acquisition by OpenAI than acquisition by Microsoft (for example)
    8. can't exclusively hang on one model / one type of model
    9. tooling around LLMs etc seems a bit boring (e.g. prompt optimisation)
  6. What kind of AI company would I start?
    1. downstream application could be fun e.g. GPT-4 colleague in Slack / Google Meet
    2. modelling is interesting, lots of interpretability challenges, bit researchy
    3. data is the most important side in the end, data enabled ChatGPT improvement
  7. Angles to a data-focused AI company
    1. getting high-quality labellers is a challenge, where are those people?
    2. rise of synthetic data, humans possibly not needed for much longer
    3. companies have so much data tied up and not much clue how to utilise it
    4. strong enterprise aspect means birds-eye view of industry at large (good strategy)
    5. dataset quality is really important, want to be confident there are no false positives etc
    6. is there a data + modelling simultaneous play? i.e. user simply uploads and selects desired end goal
    7. active learning seems really strong, makes for good demos at least
    8. possibility that massive open source models removes need for small specialised models
    9. I like the synthetic data / AI-labelled data angle