The discussion highlights that the primary bottleneck for leading AI companies like Anthropic and OpenAI is not demand, but the supply of data center capacity and power. Elon Musk's rapid construction of data centers (Colossus, MacroHard) and subsequent lease to Anthropic demonstrates a strategic pivot to becoming a key infrastructure provider in the AI ecosystem.
The speakers present staggering, albeit speculative, growth figures for Anthropic, suggesting its revenue is growing exponentially and could reach a trillion dollars by 2027. This growth is directly tied to its ability to secure more compute, positioning it as a dominant force that could potentially become 'the most powerful monopoly ever created'.
By leasing data center capacity, Elon Musk is now a direct competitor to hyperscalers like AWS, Azure, and Google Cloud. This move is framed as a brilliant strategy that monetizes his infrastructure investments, subsidizes the training of his own Grok models, and adds a significant, tangible revenue layer to SpaceX's valuation story.
There is a growing political focus on AI safety, sparked by the advanced capabilities of models like Anthropic's Mythos. The speakers discuss reports of a potential Trump administration considering a new agency to approve AI models, akin to the FDA, which they argue would stifle innovation and lead to regulatory capture by incumbents.
Keep pulling the thread on Ken Griffin.