In the B.C. years (c. <2022) the correct strategy was to eat ramen and build some weird shit like renting people's bedrooms or delivering food to ur iPod. Startups are no longer that game at all. Not even close. It's a totally different game. The game is now this: Picture the world as an isometric board, like Civ or Factorio. It seems obvious we are past the agrarian stage and before the blast off stage. The board is going to be filled with humanoids, satilites, and other crazy stuff in a few turns. So, pick an obvious Commodity or Category to supply to the world, raise enough money to be relevant in that category, and become one of a few major players in that category. Don't apply too many IQ points. Move fast. Just be a commodity. Fit into the category. Have a good brand. Be operationally excellent. Fundraise well. There's way less PMF "do people want this" risk - what remains is execution risk. Of course, that's one kind of startup. And I think that's where most of the money will be made. Another kind of startup is more like an artisan startup - it's about rebuilding existing categories with taste, like Apple building an awesome PC. But this is most relevant for Human Computer Interaction companies. If you're building a computationally verifiable company, a capital- and engineering-constrained company, not a taste- and marketing-bottlenecked company, it's best to go Elon mode.