OpenClaw: $1.3 Million in AI Tokens and the Controversy Grows

Peter Steinberger, the creator of OpenClaw, recently shared a monthly bill of $1.3 million in AI tokens, primarily for the development of his project. This impressive amount, covered by OpenAI, has sparked intense discussion about the costs associated with tokens in the field of artificial intelligence.

Steinberger posted a screenshot of his monthly bill, calculated by OpenAI, on social media. This post immediately drew attention, with some users questioning the high amount and suggesting that the budget could be better spent on hiring engineers. In response, Steinberger explained that he was testing the question: "How would we build software in the future if tokens don't matter?"

On Friday, Steinberger also shared a screenshot of his tool CodexBar, which shows token expenditures across different AI coding tools. The image revealed that his token spending on the OpenAI API amounted to nearly $20,000 for a single day. Over a 30-day period, these expenses totaled $1.3 million, a figure that shocked many observers.

Working at OpenAI, Steinberger has access to free tokens, allowing him not to pay these amounts out of pocket. He clarified that these token funds are "benefits" provided by OpenAI to support the development of OpenClaw. This situation illustrates how access to free computing resources has become a major selling point in recruiting AI talent.

The culture of "tokenmaxxing," which involves maximizing token usage, is gaining traction in Silicon Valley. OpenAI is among the companies that encourage this trend with competitive rankings based on token usage.

OpenClaw, the AI agent developed by Steinberger, has become a rapidly growing open-source product, driving the purchase of Mac Minis. Steinberger has also shared projects on GitHub, ranging from sleep mode tools to systems enabling AI agents to make phone calls.

Some users perceived Steinberger's post as a marketing strategy to encourage more spending. One user commented: "Someone is burning enough tokens to fund a small startup." In response to the criticism, Steinberger defended his choices, asserting that the automation enabled by AI makes the project extremely efficient. However, some users expressed skepticism, stating that a project spending $1.3 million in tokens per month could not be "efficient."

Despite the criticism, Steinberger maintains that his token expenditures are justified by the efficiency they bring to the project. He reaffirmed that the $1.3 million figure accurately reflects his consumption, at least on the account in question. This situation highlights the financial and strategic issues related to the intensive use of AI in technological development.