As the AI industry enters a capital-intensive investment phase, compute and infrastructure commitments often running into hundreds of billions of dollars have repeatedly sparked market discussions about bubbles and financial sustainability. According to media outlets including Jiemian News, on November 19, Anthropic CEO Dario Amodei publicly responded to AI-related investment commitments of as much as $900 billion, saying that when viewed alongside the company's current cash reserves and annual revenue scale, that figure "doesn't look quite so enormous."

Interpreting the scale of investment in light of cash reserves and revenue size

To outsiders, $900 billion is undoubtedly an astronomical sum, easily evoking aggressive capital expansion. However, Amodei offered a different interpretation from the perspective of the company's own financial scale: when this AI investment commitment is compared with Anthropic's current cash reserves and annual revenue scale, its size is not as astonishing as it appears on the surface.

Sticking to conservative principles in forecasting compute demand to guard against overestimation risks

In addition to discussing the relative scale of the funding, Amodei also explained the company's forecasting logic when planning compute. He said that when estimating future compute demand, Anthropic usually takes a relatively conservative approach and tries to avoid overestimating demand, so that its infrastructure deployment can remain more steadily aligned with the pace of actual business development.