I love reading Ed's rants, but he's too stuck on the whole revenue projection thing.
OpenAI is in a zero-sum, war of attrition game vs. competitors. It doesn't matter how much money they burn through or what funny-money estimate they don't make.
The only thing important is that they be at the top of the heap when everyone else runs out of money. Then they'll be the only game in town and can charge whatever they want. They're also not public and don't need the hassle unless the money spigot runs dry. There could also be tax advantages to showing so much loss that can accrue to them down the line.
It's the Amazon playbook. Lost money for ages. Everyone else dropped off. Now they reap the rewards.