Good article and topic. What's missing though is so critical for our time: R&D. Yes -- it's usually wise to act like investors don't exist. But are you going

Good article and topic. What's missing though is so critical for our time: R&D. Yes -- it's usually wise to act like investors don't exist. But are you going to look at the world and go "oh, what we REALLY need is another SaaS company!" The most pressing problems today require hardware, software (incl. data science), and a research component. How to fund it, if not with investors? (They don't fund R&D anyway.) Take a dual-use approach. Bootstrap with government R&D contracts -- and in parallel, commercialize it, so that you don't get stuck in gov too much of course. And/or license technology from federal labs or agencies. E.g. go to Lawrence Livermore National Lab, or Lawrence Berkeley, or Los Alamos, or the many agencies, NASA, NSF, DoE, DoD (which is colossal) and connect with tech transfer folks. They have advanced technology sitting on shelves, waiting for capable entrepreneurs to come and pick it up. Same for many academic institutions. This is how we get true technology companies like Qualcomm, and many, many smaller but by no means less significant companies. Make not just "what people want" but also what is technologically needed in this partly fallen world today.