Robert Ta notes that Anthropic, OpenAI, and SpaceX each have near-trillion-dollar valuations, with SpaceX reportedly targeting $2 trillion. Jonathan asks whether either SpaceX or OpenAI might go public soon.
Anthropic, OpenAI, SpaceX, all with near trillion-dollar valuations, or SpaceX is trying to go for a $2 trillion valuation last I heard.
Actually, anything on that, on the SpaceX and OpenAI impending IPOs?
Any perspectives on those you want to bring in, Jonathan?
One thing that's really interesting here is if you look at what the risk to any of these businesses would be, I think recursive self-improvement is both a risk and opportunity. And like the choice for Carpathia to join Anthropic and say this is the best place for me to be to bring recursive self-improvement to life is a really kind of strong insider-informed view. And if I was to say like even amongst those 3 companies who will make the best traction with recursive self-improvement, what markets will in fact, that would be the biggest wildcard I see. I think like RSI from an insider's point of view, like what is venture capital going into? I think recursive self-improvement is like maybe the bet about how LLMs will evolve or what will be beyond LLMs. And so I think the underlying assets that at least OpenAI and Anthropic have built are going to need to consider this aspect of whether recursive self-improvement is going to define their companies going forward. Is it going to impact their companies going forward, or is it just another idea, kind of like AGI, that might be like a target that we're moving towards but isn't a real thing in terms of manifestation? And then I think SpaceX is a whole other vector. They clearly have the most data center capacity.— I forget what they labeled as like a $26 trillion market. They could even not offer their own AI products and they could just offer data centers in space. So you could have these two leading AI companies relying on companies like SpaceX to create data center capacity for them in a harmonic kind of like in the same way you have NVIDIA's and Microsoft's and Google's kind of, uh, bringing the Magnificent Seven just co-evolving. Or you could see any one of them sort of maybe leapfrog each other with RSI-type innovations. It's a crazy time.
