Altman, Amodei and Musk are fighting without rules for the biggest prize in business. Never before has so much money been demanded from investors in a single year, writes The Economist
If there has been a common narrative around Artificial Intelligence this year, it is that technological advances are shaking up the global software industry. To sustain this aggressive offensive, labs on the front lines of AI must secure extraordinary amounts of funding.
This has set the stage for three giant initial public offerings (IPOs) in the space of a year: OpenAI, led by Sam Altman; Anthropic, led by Dario Amodei; and SpaceX, led by Elon Musk, which has merged with xAI, his modeling company. To make matters even more electrifying, the three hate each other.
All three are aiming for stratospheric heights. OpenAI, recently valued at $840 billion, is reportedly aiming for a valuation of $1 trillion, or 40 times its annual revenue. Anthropic, last valued at $380 billion, about 20 times its annual sales, is likely to be valued above $500 billion.
The most ambitious of all is SpaceX, which was valued at $1.25 trillion when it merged with xAI and is reportedly aiming to go public at a valuation of $1.5 trillion. That would put it among the ten most valuable companies in the world, and could even make Musk the first trillionaire in history.
Even a single IPO of this size would put pressure on the markets. Tomasz Tunguz, a venture capitalist, points out that if each company offered about 15% of its shares to the public, as is typical, the total amount raised would be roughly equal to all IPOs in America in the last decade.
This also explains the rush of leaders to stay ahead of the curve. Ego also plays a role, especially as the race to develop models has become tighter. To understand the clash between them, it helps to look at the personalities: the mercenary (Altman), the missionary (Amodei), and the visionary (Musk).
Let's start with Altman. Some admire him for starting the AI race with the launch of ChatGPT in 2022 and for the way he has outpaced rivals in securing funding. But he is also an opportunist.
This was seen in a recent clash between Anthropic and the Trump administration over restrictions on the Pentagon's use of its technology, which led to the lab being classified as a supply chain risk, a labeling the company sued to have removed on March 9 and which could jeopardize its IPO.
Unlike Musk, who openly attacks Misanthropic, Altman tried to act as a mediator between Anthropic and the Trump administration. However, OpenAI, like xAI, benefited from securing classified defense contracts at the expense of its rival.
Amodei, who left OpenAI to co-found Anthropic in 2021, takes a safety-focused approach and contrasts it with rivals' race to scale. But he's not soft.
In an internal memo during tensions with the Pentagon, he called Altman “untruthful” and opportunistic toward Trump. Anthropic’s didactic approach has also served as marketing: the Claude chatbot has gained popularity, and the company’s response has been praised even by OpenAI employees.
Musk is even more hostile to Altman. According to him, OpenAI, which has given up its non-profit status, has abandoned its mission to develop AI for the benefit of humanity.
He is demanding the return of the $38 million donation and has sued OpenAI and Microsoft for up to $134 billion. If he wins, he could bankrupt OpenAI.
Musk seems to trust no one else with control of technology. He aims to “extend the light of consciousness to the stars” and plans to send data centers into orbit to harness solar energy.
In the real world, however, money remains an issue. xAI generated about $500 million in revenue last year, far less than its rivals. A significant source is contracts with the US government.
But there are complaints that the Grok chatbot is unreliable. The limited capacity compared to rivals is one reason Musk has linked xAI to SpaceX, which has generated about $8 billion in operating profit on $15–16 billion in sales.
OpenAI, with about $13 billion in revenue, is the largest and has the broadest business model. It aims for $30 billion this year, half of which comes from subscriptions and advertising on ChatGPT (with over 900 million weekly users). The rest comes from businesses and direct use of the models.
Anthropic, with $4–5 billion in revenue, is following closely. It focuses more on corporate clients and is having success with its Claude Code tool.
By February, OpenAI had reached $25 billion in adjusted revenue, while Anthropic had more than doubled that to $19 billion. This also explains OpenAI's aggressiveness in government contracts.
Fierce competition will put pressure on prices, while data center investments increase. Neither company is close to turning a profit. OpenAI predicts $660 billion in investments by 2030. Even SpaceX needs external financing for its ambitious projects.
All three, therefore, have no choice but to turn to public financial markets. The success of IPOs will depend on investors' willingness to accept the long and uncertain path to profit.
Meanwhile, there is another alternative already listed: Alphabet, the parent company of Google. Its Gemini model is constantly improving, while the company generated $132 billion in net profit last year.
With a valuation of $3.7 trillion and a more reasonable price-to-earnings ratio, CEO Sundar Pichai could stay above this clash./monitor
