Anthropic’s highly anticipated IPO is getting a reality check. Bloomberg reports that some prospective investors are discussing a valuation of 1.2 to 1.3 trillion dollars, down from the 1.4 to 1.5 trillion floated before and well below the 2 trillion talked about in early October. Maria E. Gavilan, a physicist turned aerospace engineer turned AI strategist, explains why she sees this as price discovery rather than a warning sign. She also looks at the questions around Nvidia’s board investigation, why she thinks Anthropic is unlikely to wait for calmer markets, and what its listing will mean for other AI companies.
Get the latest news and updates on FINTECH.TV
Anthropic’s Reality Check and What the First Public Frontier AI Lab Will Mean for Markets
News about Nvidia is making headlines this morning, with its board investigating whether its CEO was targeted by a Chinese spy. What went through your mind, and how does it affect how the country looks at AI?
With the news over the weekend, we have to be mindful that there are two aspects to detach. One is the disclosure of the personal life of the CEO and that relationship. The other is the implications of the business Nvidia was doing with this person, in Singapore. For context, Nvidia had some business with the CEO of a company there, and there have been conversations around the Asian market and how Nvidia can sell more chips there. We will have to wait to hear more. Nvidia is going to do its internal investigation, which does not mean it is a legal investigation, and we will have to wait for the details and for what actions the US government might take. But again, we have to decouple the two things: the personal life of the CEO, and the alleged spy that was mentioned. We will gather more details this week and keep our eye on the news.
Does turmoil at Nvidia add another risk for Anthropic investors, given it relies on Nvidia chips?
We will have to wait and see. If we focus on Anthropic, what has happened lately, especially the conversations around valuation, is more a case of price discovery. Ahead of the IPO and the S-1, Anthropic had not disclosed its Q3 revenue, and then it recently amended the filing and disclosed it. The Q3 revenue was around 17 billion dollars, which is very encouraging. But the detail that made investors nervous is that the amended filing disclosed spending and potential capital expenditure, what is needed in costs to keep all of this compute power going and ahead of all the infrastructure that needs to be built. Those numbers were really high, and that led to reevaluating the valuation. All of this is somewhat expected, because we are in this phase of price discovery. Anthropic is the first AI company really attempting to go to market. Other companies have other segments and other markets, so they are not only AI. All of this is new to a company like Anthropic, and also new to the market.
Valuation has fallen from as high as 2 trillion dollars to 1.2 to 1.3 trillion in about a month. Is the market finding a fair price, or is this a warning sign?
It is more about the discovery process, and I anticipate it will happen with OpenAI as well. These companies come back with revenue figures and capital expenditures, and everyone makes projections of what the valuation should be. Honestly, all of these companies need some adjustment. I think some of them started really overvalued, and over time, as they disclose revenue every quarter and how much they are spending, the adjustments will reflect reality. I personally would not be as concerned, because this price discovery process is anticipated and will hopefully balance out how much real valuation we are talking about once these companies are in the market.
There are talks of a smaller fundraising target or even a delay. Should Anthropic push ahead or wait for calmer markets?
Every time we talk about calmer markets that would be ideal, but I do not think there is such a concept around AI. Volatility is very high, and the last few days are proof of that. Very likely Anthropic is not going to wait for calm waters and will continue moving forward. They are trying to be the first in the market, they need to raise significant capital for upcoming infrastructure projects, and they do not know when OpenAI might move next. OpenAI confidentially filed a draft of its S-1 and said it wanted to pause its plans to go to market, but it is not clear whether it is just waiting for Anthropic to make the next move. So I would imagine Anthropic will want to move forward, though the last few days have been so interesting that we will see how much they can delay completing the process.
This could be the first major IPO from a company building frontier AI models. What does it mean for other AI companies hoping to go public?
There has not been any other company in the market yet whose only business is AI. Companies like Microsoft, Google and even SpaceX have different segments and different sources of revenue. This is the first time we will see a public benchmark of how much investors will pay per dollar of revenue at a frontier lab. Until now, all the valuations have been set in private, by venture capitalists and strategic partners. Once these companies go public, the public markets will scrutinise revenue growth against the gigantic infrastructure investment, so the rules of the game are going to change. They will have to reevaluate how much they invest in research and development, focus on testing massive corporate adoption, and see how that translates into a profitable business. Anthropic, being the first, will set the benchmark for the companies that want to do the same next.