Bloomberg Television6 min
Claude Users Excited to Invest in Anthropic, Says Ann Berry
Interview
- Investor enthusiasm for Anthropic stems from the ability to directly observe user adoption of its Claude model, contrasting with hardware investments like semiconductor chips where usage is less tangible.
- The analyst projects Anthropic is in the "early innings" of capturing market share from Microsoft and traditional search engines.
- While Anthropic's current valuation is described as "astronomically expensive" on an EBITDA multiple, the analyst identifies an attractive long-term path to cash flow generation based on EBITDAR multiples.
- Anthropic's strategy of incorporating search data is viewed as a mechanism to increase model efficacy and product quality for enterprise clients, thereby accelerating a user adoption flywheel.
- Alphabet (Google) shares experienced significant intraday volatility, swinging from a high of +4.6% to a low of -1.5%, a movement partly attributed to investors hedging on Anthropic's pre-market performance.
- Microsoft and Alphabet are beneficiaries of private placements in Anthropic, linking their stock volatility directly to Anthropic's market sentiment.
- The long-term net impact of AI on legacy tech giants remains uncertain, with potential threats to existing business models (e.g., office suites) competing against positive drivers like increased chip consumption.
- A prevailing strategy among investors is to "spread bets" rather than predict specific winners in the AI infrastructure race.
- Concerns exist regarding the economic return on investment (ROI) for massive capital expenditures (CapEx) in AI infrastructure, as earnings calls lack concrete data on returns from these spending levels.
- Kai-Fu Lee noted that US companies often engage in wasteful spending on AI tokens despite the availability of cheaper, viable alternatives.
- National security concerns are identified as a primary barrier preventing US companies from adopting cheaper, non-US AI models.
- Chinese AI models compete globally by offering "open weight" architectures that allow enterprises to run models on-premises, addressing data sovereignty concerns common in Europe.
- While US and European markets prioritize models perceived to win or ensure data control, other global regions are prioritizing cost-efficiency, creating a potential scale advantage for Chinese models.
- The market dynamics of AI adoption currently mirror the trajectory of the electric vehicle industry, characterized by a lack of consensus on a definitive winner.