Conference Presentation, Interview
SpaceX Launches Largest Ever IPO | OpenAI Files to Go Public | Uber Cuts 23% of HR
- Elon Musk intends to fix the SpaceX IPO price at $135 per share to value the company at $1.8 trillion, bypassing traditional price discovery, though a 30% probability exists for the stock to trade down with no immediate pop and a potential medium-term valuation decline to $1 trillion from an initial $1.7 trillion.
- Over the next 12 months, the SpaceX price retention is unlikely with a significantly higher probability of a higher price eventually, while the stock is expected to begin an "inexorable rise" following major news, with an Ontario Pension Plan investment projected to yield a "magnificent return."
- OpenAI is expected to file for an IPO, though the timeline may be delayed to manage expectations, potentially triggering a "fun fall" season if the process accelerates, as the scale of such companies exceeds private funding capacity.
- AI is predicted to become a persistent "always on" fabric of life within two years, rendering non-persistent AI "archaic," while standard business practice evolves with token economics; however, IT budgets may eventually become unsustainable, forcing spend management.
- Apple is expected to deliver a compelling consumer experience using Google's models as default rather than developing its own, focusing on the consumer side which is deemed "tougher" than the enterprise space due to the requirement for "delightful experiences" over productivity.
- The Uber HR function is anticipated to see a 23% reduction partly rationalized as efficiency rather than solely AI automation, with future predictions of AI serving as a superior VP of HR, while robotaxi progress in Madrid via partners is expected to lead Uber to become the coordinating aggregator for the fleet.
- Revolut's success is forecast to pressure traditional European banks out of the market, though the fintech revolution faces limits as the largest banks fail to lend, creating a conflict in the traditional model of recycling savings.
- Founders are expected to treat VC interactions as sales and overcome fundraising rejection, while the startup sector is projected to operate at "half the size" for the same revenue, creating a divergence between lean companies like Lovable and those building large enterprise sales forces akin to Oracle or IBM within five years.
- Companies leveraging AI efficiently are expected to achieve revenue per employee of $1 million to $2 million, making such roles highly desirable, whereas the market for foundation models is predicted to consolidate into an oligopoly of four or five players within two years, threatening business models like Perplexity and Nabeus.
- Chinese open-source models may eventually go close-source, while an open-source US competitor is expected to emerge within two years to challenge frontier models; Microsoft is predicted to "grind its way to good enough" with new AI models lacking web search capabilities over the next three to four years, though it risks becoming a "sucking sound for talent and money" if it fails to control foundational technology.
- Bending Spoons is expected to grow via acquisitions and cost-cutting until it exhausts "affordable targets," while Databricks is anticipated to delay its public offering as private rounds suffice for capital needs, despite a long-term expectation that cheaper public capital will make an IPO logical.
- Venture capital return expectations are rising to 7 to 8x routinely, making $5 to $8 billion exits difficult for large funds, and growth narratives tied to "one to 100 in a year" are seen as limited by GDP and human nature, meaning stocks priced for perfection will filter quickly upon any miss.
- Replit is noted as hiring sales reps while others like Lovable remain lean, and Elron Musk's acquisition of Cursor is predicted to be a "prescient buy" given the target of $6 billion in revenue, with the capital markets described as "risk-on" as long as investors remain brave.