| We're less than a month away from our Financing the AI Revolution summit. Join me on the afternoon of April 28 for debates and discussions with investment leaders from General Catalyst, KKR, Benchmark and Fidelity about the future of data centers, chips and AI apps. You can find more details here and purchase tickets here. If you're interested in a sponsorship, please reach out to sponsorships@theinformation.com. We hope to see you there!
Welcome back! OpenAI CEO Sam Altman described the awe-inspiring usage of ChatGPT's new image generation tool a few days ago as "biblical." His use of religious terminology fits the moment. The money that Altman, Elon Musk and tech giants require to finance chips, data centers and power is a test of faith for Wall Street. You see it in OpenAI's $40 billion funding round, the largest ever by far for a startup. The money came from SoftBank, the ultimate true believer. "Biblical" is also a good description for the $35 billion OpenAI has told investors it expects to burn over the next few years, as we reported. SoftBank's Masayoshi Son is largely footing the bill. He will need to convince more investors and bankers that it's worth the risk if he is going to borrow enough money to finance the investment in the company and its data-center expansion. You see it in Elon Musk's head-scratching merger of X and xAI. You need a near-religious conviction around both Musk and AI to buy the $110 billion combined valuation. And you need to ignore the strange reality that advisers like Morgan Stanley played both sides of the deal, the WSJ reported. But the transaction makes sense if you believe it will help deliver to Musk the ungodly number of chips he will need to compete with Altman. These pivotal financial moments in the AI race have come at a time when Wall Street seems to have some doubts about who's going to make money in the end. You see those concerns in the reception to CoreWeave, which is trying to become the "AI hyperscaler," with its own biblical cash burn. The big debate on Wall Street during the IPO centered on a seemingly bland accounting question that is, at its core, a question of belief. The boring question at play was: How long should it take for the chips CoreWeave purchased to depreciate? The company said six years. Investors didn't have the same faith. The answer would be crucial to making the investment work. The company downsized the deal. And I learned that investor sentiment had turned so sour that CoreWeave's bankers needed to allocate about half of the $1.5 billion it raised to just three investors, putting much more stock in fewer hands than in typical IPOs. One of the offering's largest investors, Nvidia, is the leader of the AI church. It also sells chips to CoreWeave and is one of its largest customers. It's on all sides of the trade. It's possible, of course, that this framing is all wrong. Wall Street, which has sold off its index of semiconductor stocks by nearly one-fifth since the start of the year, is the one acting irrationally, the converts say. Gavin Baker, an Nvidia and Musk investor who runs Atreides Management, wrote on X today the disconnect between Wall Street bearishness on semiconductor stocks was "wild" given AI companies' professed need for chips and data centers. AI believers would claim that they're the ones with rationality on their side—that money needs to keep flowing into companies selling chips and building data centers because customer demand underpins the investment. Look at OpenAI's 30% revenue growth in three months. Listen to Microsoft chief technology officer Kevin Scott saying this week that 95% of new code would be AI-generated within five years. Skeptics may see the light again. And at least on Tuesday, investors were convinced in one way. CoreWeave's stock closed the day up 42%. CoreWeave cofounder Brannin McBee brushed off the bubble talk to me on Friday. "This conversation around an AI bubble seems to come up every three to six months or so. And then it drops away because people see, 'Oh my god, all these companies are continuing to grow so much.''' What do you believe? | 
0 comentários:
Postar um comentário