Pre-IPO, Anthropic Shows Up OpenAI with Surprise Profit Projection

What’s worth the most, brains or bucks?

OpenAI this week claimed that one of its research models cracked the “Erdős planar unit distance problem,” a challenge that has bedeviled human eggheads for 80 years. Not to be outdone, rival Anthropic recently told investors it had solved the seemingly equally intractable arithmetic problem of running an AI company where revenue tops costs, claiming it was on track to post a profit in the second quarter, according to a report in The Wall Street Journal.

As both companies prep IPO filings, with OpenAI’s coming as soon as today, we’re guessing investors might be a little more infatuated with the latter’s eureka moment than the former’s. So how exactly did Anthropic pull it off?

A Wrinkle in Time

Previously, Anthropic told investors it probably wouldn’t achieve full-year profitability until 2028 at the earliest, though the meteoric growth rate of its Claude AI tools is helping it reach that milestone sooner on a quarterly basis. The result? An expected $559 million in operating profit on $10.9 billion in revenue in the current quarter, up from $4.8 billion in the first quarter. 

A deeper look at the equation reveals a few more key variables at play: 

• First: Efficiency. Per financial projections seen by the WSJ, Anthropic expects to spend 56 cents on compute power for every dollar generated, down from 71 cents in the first quarter. It’s a sign that the AI industry is crawling toward the efficiency of scale that built Silicon Valley.
• The second and likely more critical variable: timing. Anthropic has been a bit slower than OpenAI to sign on to massive data center deals, allowing it to tip into profitability … for the moment. User demand is already pushing up against the limits of its compute capacity, and the company told investors that it likely won’t be profitable for the whole year as it loosens its grip on the capex hose. Scale still has its costs.

Open the Books: Those costs, however, are lower than OpenAI’s. The competing AI firm and ChatGPT-maker actually posted more revenue than Anthropic in the first quarter, scoring around $5.7 billion in sales, according to a report by The Information this week. But the company doesn’t expect to achieve profitability until around 2030. Its compute demand is still dominated by consumer users on free tiers, whereas the Anthropic user base is largely composed of enterprise clientele. OpenAI’s models are also tightly intertwined with Nvidia’s architecture, which is proving more expensive than the Google TPUs and Amazon Trainium chips that Anthropic has largely leaned on.