Michael Burry Questions AI Agent Traffic Business Model
Investor Michael Burry responded to Elon Musk's prediction that AI agents will far exceed human internet traffic usage, stating that this represents a "lower threshold than most people imagine." He raised a core question: it remains unclear who will pay for the social activities of AI agents. His remarks reinforce the narrative that the growth of AI traffic is disconnected from sustainable monetization, with funding and attention shifting towards AI applications that can demonstrate willingness to pay and a closed business loop, putting pressure on narratives driven by pure traffic expectations.
Source: Public Information
ABAB AI Insight
Michael Burry is known for accurately predicting the 2008 crisis and has recently expressed caution regarding AI infrastructure and valuations, emphasizing the mismatch between capital expenditure and actual returns. His comments focus on the lack of a closed business loop: acknowledging that agent traffic may surge, but questioning who bears the costs of social interactions, aiming to shift the discussion from technological possibilities back to fundamental economic questions—who pays and how to monetize.
Similar cases can be seen in the traffic bubble before the early internet advertising model took shape, as well as the current tension between declining AI inference costs and users' willingness to pay. The industry is in a phase where traffic predictions are high but monetization paths are unclear.
Essentially, this is about the transfer of pricing power: traffic itself no longer automatically converts to revenue, as the marginal costs of agent behavior need to be clearly assigned; otherwise, high traffic merely becomes a burden on infrastructure rather than a source of value.
ABAB News · Law of Cognition
- Traffic growth is a low threshold; willingness to pay is a high threshold.
- Who pays for agent social interactions determines whether the AI economy can close the loop.
- Predictions are easy; finding the real payers is much harder.