Even before ChatGPT you had things like Context Free Grammars [1] that could give results kind of like GPT, superficially. You could certainly get it to generate a bunch of sentences that could be used like business ideas. That's kind of what I was going for.
> powered vertical landing to reuse rockets
Most of your points are fair, but I reject that this is actually a clever thing on Musk's end. NASA experimented with reusable rockets in the 60's. The X-15, for example, while slightly different, demonstrates that this was something being toyed with quite awhile ago.
No doubt, SpaceX's achievement here is notable and cool, and to whatever extent Elon helped with that he deserves credit. But I think "having the idea" does not really count.
Indeed.
But between this and what you say of CFG generators, I think you misunderstood me.
He didn't just have the idea, he got it to work.
If it was just ideas that counted, I've got one for 3D printing diamond with CVD, and metal with electro-deposition, and a way to radically improve Farnsworth-type fusion.
You'll note I'm not even the most famous person with my own name, let alone for any of those things.
I guess what I was trying to say in regards to a reusable rocket is that it wasn't an idea that people laughed at. It's an idea that's pretty obvious: rockets are expensive, the way that the previous rocket stuff had worked is pretty wasteful, wouldn't it better if we can reuse more?
So hiring a bunch of engineers to build an obvious idea just doesn't impress me that much. The engineering impresses me, no doubt some very clever did something very cool, but I'm just not 100% sold that Elon deserves the lion's share of the credit on this.
I've played with Markov chains, and… no. While you can totally generate some stuff with them, they're not even close to the level of ChatGPT — and ChatGPT is the reference point I use these days for "are you sure you know about this subject?", in part due to someone else describing it as "mansplaining as a service". For a lot of the stuff Musk suggests which horrifies domain experts (like how the tunnels are so narrow that if a car catches fire the people in it can't get out), what he says is stuff I'd put in this category.
But he seems to have more lucid and self-aware moments besides those, and many of which are also things which are more mocked than being horrifying (e.g. powered vertical landing to reuse rockets). It comes across to me (not a psychologist) as something of a bipolar personality — or as if some of his ideas resulted from snorting a white powdered stimulant (a category which includes caffeine, unwise as that is).