Leading artificial intelligence company OpenAI is expected to lose up to $5 billion as a result of serious financial instability.
An analysis by The Information indicates that OpenAI exceeds its rivals in the generative AI sector on operating expenses. According to the estimate, this year’s costs for OpenAI’s inference and training might reach $7 billion, in addition to an additional $1.5 billion for staff. This stands in contrast to Anthropic, another AI player who is anticipated to spend $2.7 billion.
The rising expenses of sustaining OpenAI’s key services, such as ChatGPT, have become a common topic of conversation during the last eighteen months. It will cost roughly $700,000 ($694,444) per day to maintain ChatGPT in operation in 2023.
OpenAI’s revenue is only about $3.5 billion despite these massive expenses, which means that its profit margins are unsustainable and industry spectators are questioning the company’s business strategy. Thanks to a relationship with the computer giant Microsoft, OpenAI has subsidized access to Azure cloud services. Over the past two years, Microsoft has also spent billions in OpenAI; yet, doubts over the company’s long-term sustainability have only grown.
OpenAI’s financial difficulties are a symbol of larger problems facing the AI sector. Concerns over the return on investment (ROI) of generative AI technology are growing among stakeholders. According to a research by software company Ardoq, senior tech executives frequently feel as though they are “finger in the air” while evaluating ROI on such technology. In the first year, just one-third of the organizations claimed to have seen a noticeable return on investment.
The CEO of Omnilndex, Simon Bain, noted that IT executives are beginning to realize that AI’s “jack of all trades approach has failed.” “Although the eye-catching demonstrations and captivating chat initially garnered attention and free users, they haven’t offered many (if any) actual business solutions,” Bain remarked. “People have therefore seen no need to pay for it.”
The difficulty in deriving value from generative AI for enterprise users, coupled with mounting costs, has led to increasing caution. Dom Couldwell, Head of Field Engineering at DataStax, noted the challenges in transitioning AI projects from concept to production, emphasizing the “hard work” involved.
“Until people … actually get to production and the business side actually sees the value, we’re going to have this gap, and no one knows quite how long that gap’s going to last,” he said.
These thoughts were mirrored by Mark Rodseth, VP for technology, EMEA at CI&T, who emphasized the constant requirement to prove AI’s worth. “Because AI is still in its early stages, proving ROI can be challenging – both to external and internal stakeholders. But this doesn’t mean companies should stop embracing AI,” he noted.