Scott Belsky: "a few open questions/thoughts from taking off the rose-colored glasses on “generative AI": what disruptions are overblown, the perils of building start-ups (or features) on others’ models, and what will (and will never) change, in no particular order:" / Twitter