On February 12, the "New York Times" published an article on Wednesday saying that OpenAI's biggest challenge at present is to turn AI into a "cash cow." The company hopes to quadruple revenue in the next year as it plans to invest tens of billions of dollars. However, time is of the essence.


OpenAI San Francisco Office

last resort

Two years ago, during an appearance at Harvard University, OpenAI CEO Sam Altman said he hated the idea of ​​showing ads within ChatGPT.

He said at the time that if ChatGPT inserted paid ads while answering questions, people would start to lose trust in the company's flagship product."I kind of see advertising as a last resort business model for us."Altman said.

However, just this week, his company started running ads in ChatGPT.

As OpenAI invests tens of billions of dollars in the raw computing power needed to build and deploy AI technologies like ChatGPT, the San Francisco startup is rushing to find new ways to generate revenue from these technologies to eventually break even. Adding ads to its chatbot is just one of many ambitious plans to increase revenue, each of which comes with huge challenges.

Financial pressure is great

Financial stress is imminent. OpenAI had revenue of about $13 billion last year, according to a person familiar with the matter. But the company expects about $100 billion in additional spending over the next four years.

Altman and his deputies have had great success raising money in recent years. However, there are only a handful of places on the planet willing and able to provide the billions of dollars OpenAI needs to purchase raw computing power.

One option is to go public on Wall Street. But even OpenAI executives admit privately that before that goal can be achieved, losses must be stemmed. In order to quadruple its revenue this year, OpenAI must tackle many things it has no experience with. It has never run an advertising business, which could hurt the value of its chatbot, or worse: lose users. It also plans to make more money by selling technology to businesses, although it already faces a slew of competitors in this area.


ultraman

Google has decades of experience selling products to enterprises. So does Microsoft. As a rival of OpenAI, the startup Anthropic continues to make progress in the field of AI programming, which may be the most noteworthy area in this emerging market.

OpenAI also proposes new business models that could drive away customers. The company recently said it wants to take a portion of the revenue from scientific discoveries made using its AI tools. Although the company later explained that the move would only affect big pharmaceutical companies, the idea has unsettled many independent scientists who use its technology.

Too much needs to be done

“OpenAI is trying to win over consumers, trying to keep up with Anthropic’s programming tools, trying to build data centers, trying to raise more money. There are just so many things that it needs to do at the same time,” said Brian O’Kelley, CEO and co-founder of Scope3, an internet advertising company. He's been in the field for two decades. "Does it really do a good job at advertising? Does it do a good job at everything it's trying to do?"

Last week, the Wall Street Journal reported that OpenAI could go public as early as December, surprising some OpenAI executives, according to two people familiar with the matter. Their main concern is that companies aren't ready yet.

As of the end of last year, about 60% of OpenAI's revenue came from consumer products and 40% from enterprise technology. Most consumer revenue comes from subscriptions: Of the 800 million users of ChatGPT, about 6% pay at least $20 per month to use more advanced versions of the chatbot. The move into advertising is aimed at generating additional revenue from the free version of ChatGPT.

Many online advertising industry veterans believe that AI chatbots like ChatGPT can eventually bring in billions of dollars in annual advertising revenue, but that may take years of experimentation. While OpenAI is experimenting, it will face competition from Google and other established advertising companies.

OpenAI has begun building an ad sales team, but the effort is still in its early stages.

"OpenAI actually doesn't have a real sales team yet," said Mark Zagorski, CEO of DoubleVerify, which works with Google and many industry advertising companies. "They need to build not only the sales infrastructure, but also the technical infrastructure needed to run the advertising business."

Enter advertising

In May last year, Altman hired Fidji Simo, a former senior executive at Facebook, as CEO of the OpenAI application business. This is a newly created position that oversees the company's numerous product lines. Seamus previously served as CEO of grocery delivery company Instacart, where she pushed the company toward an advertising-based business model.

Over the next few months, OpenAI poached hundreds of employees from X and Facebook parent company Meta, many of whom had worked on advertising products.

DoubleVerify CEO Zagorski compared OpenAI to Netflix, which he noted took two years to build a viable advertising business. During this period, Netflix outsourced a lot of work to more experienced companies.


Ultraman recruits Simo to promote advertising business

Even as OpenAI begins to get involved in the advertising business, the company still hopes to increase the proportion of revenue from enterprise products (technology for enterprises, government agencies and other large organizations) to 50% by the end of this year.

"This is a core issue for technology investors right now," said Karl Keirstead, an analyst at investment bank UBS. "OpenAI has no choice but to move more aggressively into enterprise software."

enterprise market

Currently, companies need to pay OpenAI to use tools such as Codex, which helps software developers write computer code, and ChatGPT Enterprise, which is designed for general office scenarios. Such tools are widely used among Silicon Valley techies, with some users paying as much as $200 a month to use them.

But Kiersted pointed out that ordinary companies may not be willing to pay such a high price for office software. At the same time, OpenAI faces increasingly fierce competition in the enterprise market, with the most prominent competitor being Anthropic and its code generation tool Claude Code. While OpenAI strives to accelerate revenue growth on both consumer and enterprise products, Anthropic is primarily focused on enterprise tools.

Anthropic recently launched a Super Bowl ad that satirizes OpenAI’s attempt to introduce advertising into ChatGPT.The tagline reads: "Advertising is about to take AI by storm. But not Claude."

Altman fired back in a post on

value share

Last month, at the annual World Economic Forum in Davos, Switzerland, OpenAI CFO Sarah Friar discussed another new way the company hopes to generate revenue. She calls it “value sharing.” For example, if OpenAI's technology helps discover a new drug, OpenAI might get a share of the profits.


Freer

A few days later, OpenAI launched Prism, a product for scientists.Many researchers, aware of Fryer's previous remarks, were concerned about whether OpenAI planned to take a cut of their scientific discoveries.

Concerned that the move was alienating customers, OpenAI executives discussed how to respond to the burgeoning controversy, ultimately deciding to focus on posting clarifying messages on social media.

Kevin Weil, OpenAI's new head of science operations, explained in a post that the company does not take a cut of the discoveries made by independent scientists using Prism. Other OpenAI executives have expressed similar stances on the X platform. But Weill hasn't completely ruled out the possibility of forming partnerships with big pharmaceutical companies and taking a cut of the profits.

Altman expressed a similar view this week at an event in Silicon Valley: "We may explore a cooperation model where we bear the costs and share the benefits."