Keep pulling the thread on Mark Cuban.
A potential market bubble could destroy many VCs, funds, and private equity firms because they are heavily invested, though it is unlikely to impact the general public.
If a breakthrough in AI price-performance significantly minimizes power requirements, many newly constructed data centers could become obsolete.
The fact that AI companies like Microsoft, Anthropic, and OpenAI need to hire thousands of "forward deployed engineers" indicates that AI technology is difficult to implement.
The platform Lovable is being used to create 770,000 applications per week.
AI agents experience "drift" and can break over time as the underlying large language models they are built upon are updated.
Using Lovable, two or three employees built internal software that would have otherwise cost an estimated $2-3 million per year to develop with an outsourced firm.
Large language models will be the most important technology for reducing political information asymmetry because their business model requires them to be truth-seeking to maintain user trust.
The valuations of NBA teams are primarily driven by subscriptions to streaming services, not by game attendance or team performance.
The current market environment is different from the dot-com bubble, as companies going public today generally have revenue and traffic, unlike the zero-revenue companies of that era.
Angel investment valuations for pre-launch companies have risen from a historical range of $5-10 million to a more recent range of $40-60 million.
Large technology companies like Google and Meta are spending all their cash flow on CapEx and are borrowing billions of dollars on top of that.
AI companies should pursue smaller IPOs ($50-100 million) to gain a public currency for acquiring legacy businesses or companies with valuable data and domain expertise.