|
Thanks for reading The Briefing, our nightly column where we break down the day’s news. If you like what you see, I encourage you to subscribe to our reporting here.
Greetings! There were plenty of surprising revelations about Meta Platforms this week, thanks to its antitrust trial, most obviously that CEO Mark Zuckerberg mused in 2018 whether to spin off Instagram. But even more striking was The Information’s report on Thursday night that Meta for much of the past year was trying to get other big tech companies to help fund its Llama AI models. It seems there is a limit to Zuckerberg’s ability to blow billions on new technology—who knew? We shouldn’t be surprised that Zuckerberg was looking around for donations to the cause (it’s unclear whether this effort is still underway). The spending on artificial intelligence development is mounting even as uncertainty about the potential return persists. We’ve seen others come up with creative financing ideas, most notably OpenAI and its Stargate group, which plans to finance up to $500 billion in AI data centers. Perhaps we shouldn’t assume Meta is the only big company asking for help from its friends. We often think of these big tech companies as having enormously deep pockets, but AI development is tearing a hole in those pockets. Meta generated $91.3 billion in cash from operations last year, out of which it spent $39 billion on capital expenditures and $35 billion in stock buybacks and dividends. This year Meta has projected capex to rise 60% to between $60 billion and $65 billion. We shouldn’t assume its operational cash flow will rise 60%, so you can see where a problem might arise. Cutting buybacks wouldn’t thrill shareholders, which are surely already wondering about the value of the AI investment. What about the other companies? Amazon, Microsoft and Alphabet reported operational cash flow ranging from $116 billion and $125 billion in their 2024 fiscal years—more than Meta, but not that much more. Alphabet last year spent almost all of its cash flow on capex, dividends and buybacks, and it is forecasting that capex spending will rise to $75 billion this year. Microsoft and Amazon have a little more cushion between their cash flow and their spending, but that cushion may wear out quickly as capex rises. All these companies (including Meta) have plentiful cash in the bank, of course, but that could diminish quickly as AI spending increases. Meta, for instance, has discussed building a $200 billion data center campus. Borrowing money to fund AI is an option. Bloomberg reported a few weeks ago that Meta was working with Apollo on a $35 billion debt financing for AI data centers. There’s also the possibility of Stargate-like consortia—Microsoft is working with Blackstone, Nvidia and xAI on a data center fund. But that’s a $30 billion effort, which the group says will eventually rise to $100 billion, which may get spent quickly. All of this begs what may be the most important question: Why is each of these companies developing different AI models? How many do we need? If short sellers needed any more reason to take a position against Tesla, they got it with our scoop this week about how Elon Musk made his Robotaxi push at Tesla despite his staff’s internal analysis showing the Robotaxi might generate relatively meager sales and might never make money. Going one layer deeper, the piece suggests Elon Musk has lost interest in the relatively mundane work of making cars in favor of the more exciting—but higher-risk—robotaxi effort. The piece is definitely worth a read. It was a big week for antitrust action, what with Google’s second loss, this time in the ad tech case brought by the government, and the opening week of the Federal Trade Commission’s trial in its lawsuit against Meta. Coverage on Meta is here, here and here. Meanwhile, as the two digital ad giants fight legal action, one of their small rivals, Pinterest, has been looking at acquisitions to bolster its digital ad business. On the AI front, we revealed how OpenAI’s latest reasoning models could come up with novel ideas for how to “tackle problems such as designing or discovering new types of materials or drugs.” We dug into how Google is trying to figure out how its search engine will evolve with AI. Big companies such as PayPal and Shopify are beginning to replace jobs people used to do with AI. Here’s a rundown of the most common and least common uses of AI. On venture funding, some U.S. creator economy startups are raising more money. Plus, investors are pouring money into startups selling healthy foods. Meanwhile, the IPO market might be dead, but it won’t stay that way forever—Figma this week began the formal process by filing IPO paperwork confidentially with regulators. Start your Easter weekend reading off with this excellent deep dive into how game engine Unity nearly blew up its business—and how it’s trying to come back. • Netflix co-CEOs Ted Sarandos and Greg Peters each earned more than $60 million for 2024 as a result of the company’s blockbuster year, a sizable increase on 2023. However, much of that is due to grants of restricted stock, which vest over multiple years and in some cases depend on Netflix’s stock outperforming the overall stock market. Introducing: Applied AI. This new newsletter explores how businesses and leaders are using AI to innovate, improve efficiency, and foster collaboration. Stay ahead with insights and stories on the transformative power of AI. Sign up here. | 
