| Over 300 organizations rely on The Information for exclusive insights into public and private companies, including in-depth analysis of the ways in which tech moves markets. Click here to contact our corporate and enterprise team to learn more. Welcome back! My ears perked up Monday when I heard the words "unparalleled capital implosion." That came from venture capitalist Eric Vishria, who unspooled some of his anxieties about the current artificial intelligence funding boom at our "Financing the AI Revolution" conference. For more than a decade, Vishria has helped VC firm Benchmark navigate the latest tech booms and busts. He drew a contrast between the funding boom for AI and the zero-interest-rate startup bonanza in 2021. This investing moment "is a lot more dangerous," he said, because AI startups' cash needs to go to chips and other infrastructure. In earlier days, money from investors would be spent on employees, which could be cut if the business faltered. Stunning revenue growth might also be illusory if customers are just experimenting with the technology. That's in contrast to software companies, where "if you got a company to $50 million to $100 million [in revenue], for the most part it wasn't going away," he said. "I think one of the harder parts about investing right now in AI companies is the variance is much, much higher in terms of outcome," he told attendees at the New York Stock Exchange. Those comments amounted to as bearish as speakers got. More skepticism would be too far outside the mainstream in Silicon Valley and Wall Street circles. Even Vishria thinks we're "going to see a couple trillion-dollar companies come out of this." In another example of how much investors are falling over themselves to back top AI founders, check out my colleagues' scoop today on the unusually outsize control ex-OpenAI chief technology officer Mira Murati is slated to get over her new startup's board. Still, it was clear most investors are trying to understand how this funding cycle could blow up in their faces. The boom relies on fast-growing digital firms, financed by venture capitalists with few downside protections. That's standard operating procedure for tech. What's new is that these companies need space in data centers, which cost a lot. For these, the funding is coming from lenders, who are paid not to lose money. You could hear that divide in comments between two different types of investors on Monday. I asked Gavin Baker, a backer of xAI who runs the hedge fund Atreides Management, how Elon Musk's AI firm would amass the computing power it needs. He said there would be "real leverage" on the company, and pointed out that the Nvidia graphics processing units xAI would buy were "very financeable…that's just a fact." That fact didn't seem as clear for Waldemar Szlezak, global head of digital infrastructure at KKR, which invests in data centers. He was asked whether KKR would back "neo-cloud" businesses that are raising debt backed by GPUs. He wasn't keen on the notion. "It's too difficult to say what is the useful life of GPUs," he said, adding: "It's probably higher on the risk curve." While venture capitalists like Vishria are voicing their worries, the investment decisions by lenders are likely what matters more to AI right now. I asked Tony Kim, who runs a public and private tech investing group at BlackRock, whether he was worried about a bubble in AI infrastructure. "I have lots of worries and fears," he said dryly. He's paying close attention to investor sentiment for funding these giant projects. "I think the market will vote. And you might see it. If you can't raise $100 billion of debt to finance the next massive data center, you'll start to see it there." | 
0 comentários:
Postar um comentário