OpenAI is strategically shifting its focus from consumer offerings to business-oriented products, aiming for profitability, as evidenced by its chief financial officer, Sarah Friar, utilizing ChatGPT for both personal tasks like finding recipes and professional duties such as summarizing emails and Slack messages. This pivot is crucial as the company grapples with the high costs associated with powering its AI systems, necessitating a strong emphasis on acquiring big business customers to offset these expenses.

The company is preparing to launch a new artificial intelligence model, codenamed Spud, designed specifically for “high-value professional work.” Friar expressed excitement about this upcoming model, touting its stronger reasoning, improved understanding of intent and dependencies, better follow-through, and more reliable output in production. This move comes amid heightened competition, particularly with rival Anthropic, in attracting corporate clients to integrate AI assistants into their workplaces.

While OpenAI’s core ChatGPT product boasts over 900 million weekly users, Friar notes that approximately 95% of them do not pay for the service. This reliance on free users strains computing resources and underscores the urgency to secure paying business customers. Both OpenAI, valued at $852 billion, and Anthropic, valued at $380 billion, are currently operating at a loss, fueling a fierce competition to generate revenue as they eye future public listings on Wall Street.

This drive for improved performance and sales of business-oriented products has led OpenAI to discontinue some consumer initiatives, such as the AI video generator app Sora. Friar acknowledged this decision was “a little heartbreaking” but necessary to ensure sufficient computing power for their new, more advanced model. This sharper focus aligns with recent statements from OpenAI co-founder and CEO Sam Altman, who emphasized the importance of narrowing the company’s objectives, a sentiment Friar echoes regarding the pitfalls of over-diversification for growing tech companies.

The strategic shift is further solidified by the hiring of Denise Dresser, former Slack CEO, as OpenAI’s first chief revenue officer three months ago. Dresser is intensely focused on engaging with corporate leaders, positioning OpenAI as the premier platform for businesses looking to automate various computer-based tasks using AI agents. She observes that companies have moved beyond the experimentation phase and are now using AI for “real work,” recognizing it as a transformative shift.

However, companies have alternatives, notably Anthropic’s Claude, which is widely adopted by software professionals. Founded in 2021 by former OpenAI leaders prioritizing AI safety, Anthropic has cultivated an image as a more responsible AI vendor. Despite a past incident involving the Trump administration and a contract dispute, Anthropic’s annualized revenues reportedly hit $30 billion, though revenue measurement differences exist between the two companies, particularly regarding shared revenue with cloud providers like Amazon and Google.

The competition remains tight, with Luke Emberson of Epoch AI suggesting Anthropic might be growing faster. In a memo to OpenAI employees, first reported by The Verge, Dresser acknowledged Anthropic’s “early wedge” in coding but expressed confidence in OpenAI’s “real structural advantage” as AI use expands beyond software developers and computing capacity increases. Dresser’s memo contrasted Anthropic’s message, which she described as “built on fear, restriction, and the idea that a small group of elites should control AI,” with OpenAI’s “positive message” of building powerful systems with safeguards, expanding access, and empowering people.

Skeptics, including AI critic Ed Zitron, raise concerns about the financial viability of AI products like ChatGPT and Claude. Zitron warns of a potential “subprime AI crisis,” where businesses reliant on these tools could face difficulties as companies try to cut costs, exemplified by Anthropic’s imposition of rate limits on heavy users. The high cost of AI technology, particularly the electricity-hungry AI computers, remains a significant challenge. Zitron cautions that even public companies can fail, especially those dependent on hundreds of billions of dollars annually just “to keep breathing.” Both AI leaders and critics agree on the expensive nature of this technology, with its long-term economic worth still under scrutiny.