I think PG was correct here about Oracle, as it leaves the door open to being a much bigger product rather than being too niche, which ironically sounds like th

I think PG was correct here about Oracle, as it leaves the door open to being a much bigger product rather than being too niche, which ironically sounds like the reason the company failed. You can imagine lots of situations in deployment where having a monitoring “source of judgement” about running systems could be developed. Not just security, but say performance, efficiency, cost. It’s not explained in the essay why the author was so certain PG was wrong. If PG hated “Reddit” and the mascot, I would say he was correct again. Reddit could have been a much bigger idea, but the branding excludes a lot of demographics and types of content. It helps to make Reddit what it is, but also what it is not. That can be helpful, but in the startup game you want to be biased towards generality: “Google”, “Apple” not “websearchengine” or “foundit” or “homecomputerco” or “appyphone.” It’s a bit like being a good coder: you develop a sense of test for elegance, simplicity, leaving the door open to expansion… and it sounds like PG has (unsurprisingly) refined taste in naming.