Anthropic's annualized revenue run rate has surged to $30 billion, a 3x increase since late 2023, potentially surpassing OpenAI. This unprecedented growth, annualized at 9700%, is driven by enterprise customers, with those spending over $1M annually doubling in just two months, signaling a major shift in the AI market.
OpenAI and Anthropic are presenting a version of profitability that excludes their largest expenses: model training and inference. Critics compare this to an airline ignoring the cost of its jets. This financial engineering aims to create a positive outlook for potential IPOs but masks their true cash burn rate.
The competition for AI dominance has moved beyond chips to securing massive energy and infrastructure. Anthropic's new deal with Google for 3.5 gigawatts of power capacity highlights this shift. This single deal effectively created a multi-billion dollar business for Google, reframing the AI race as a battle for power plants.
At companies like Meta, a new practice called "token maxing" is being used to measure productivity, where engineers compete on leaderboards to consume the most AI tokens. Promoted by leaders from Nvidia and Meta, this metric is criticized for being easily gamed and not necessarily reflecting true productivity.
A significant credibility gap is forming between AI executives' talk of "superintelligence" and the often buggy, frustrating reality of using current models. This disconnect devalues serious policy discussions and creates cynicism, with observers noting we are in an "extremely capable tool era," not a "new social contract era."
OpenAI's recent policy paper suggests societal solutions like a public wealth fund and higher capital taxes. However, it's being heavily criticized for its noticeable lack of any commitment from OpenAI to fund these initiatives or voluntarily adopt the policies it recommends, making the proposals appear hollow.
A contrarian view argues that encouraging high token usage ("token maxing") is a valid short-term strategy. The rationale is that the engineering challenge of building systems capable of consuming tokens at massive scale is a significant achievement and a proxy for deep AI integration, making the raw cost secondary.
The AI industry's public communication strategy, which heavily emphasizes risks and downplays tangible benefits, is backfiring. By constantly validating fears without clearly articulating a positive vision, AI leaders are inadvertently encouraging public skepticism and making people question why the technology should exist at all.
Polls show a majority of Americans now believe AI will do more harm than good, an 11-point jump in one year. This negative sentiment is growing despite, and perhaps because of, rising adoption. The paradox is that increased AI fluency correlates with decreased optimism, particularly about the job market.