SpaceX stock is acting like one of the company’s rockets—soaring into orbit and coming down just as steeply. But Elon Musk’s space company is just the most visible sign of a market struggling to value technology companies and their expensive race into an AI-powered future. 

SpaceX lost more than $400 billion in market value on Monday. Although the immediate trigger appeared to be its inaugural investment-grade bond sale, the drop is a sign of the underlying issues. Investors don’t know the appropriate valuation for the company’s dream of putting artificial intelligence into space. 

It’s not a problem unique to SpaceX. Google-parent Alphabet suffered its worst ever one-day loss in market value Monday, down $225 billion. The general tech selloff was exacerbated by news that research scientist John Jumper—a Nobel Prize winner—was leaving Google DeepMind to join AI start-up Anthropic. Valuations of existing Big Tech companies are complicated by the prospect of Anthropic and ChatGPT-developer OpenAI coming onto the public market shortly. 

So far, the saving grace for the AI trade has been semiconductor stocks. The value of companies such as memory-chip maker Micron have soared on the basis that they are the clear beneficiaries of the flood of AI spending. But even that faces its complications—how long can Micron really sustain gross margins above 80% when Chinese competition looms? Expect plenty of questions on that when Micron reports earnings after the market closes Wednesday. 

Investors have been willing to buy the dip on any tech slump up until now. But it would only take one of the big players to signal they are reining in their capital spending to change that. Keep an eye on Microsoft, where CEO Satya Nadella has been increasingly critical of the high costs of leading-edge AI models. 

SpaceX might be at the extreme edge of both tech ambition and volatility but don’t assume its tumble is a one-off case.

—Adam Clark

Get more of the journalism you love. Choose Barron’s as a preferred source in Google.

Competition for top AI talent has the potential to clobber even the biggest Big Tech stocks, as Alphabet experienced on Monday, the first trading day after John Jumper, a senior research scientist and Nobel Prize winner, said he was leaving Google DeepMind for artificial-intelligence start-up Anthropic.

• The Google parent’s stock ended the day down 5%, shaving $225 billion from its market value, the biggest drop for the company ever, according to Dow Jones Market Data. Jumper is leaving after nine years. DeepMind is the backbone of Google’s advanced AI models.

• Google praised Jumper’s “significant contributions” to DeepMind’s work and wished him well. Anthropic confirmed to Barron’s that Jumper will be joining it. The AI start-up declined to comment on how it is prioritizing talent acquisition or whether Anthropic plans to make any more high-level hires in the near future.

• Jumper helped create AlphaFold, DeepMind’s AI system that predicts protein structures from their amino acid sequences. His departure comes shortly after Noam Shazeer, a vice president of engineering at Google and a key member of the Google Gemini team, said he was leaving for OpenAI.

• Big Tech along with AI start-ups Anthropic and OpenAI are currently fighting for the top AI talent in the marketplace as they all vie for superiority. As a result, pay packages have stretched to the hundreds of millions, and big-money acquisitions of smaller rivals have become commonplace. 

What’s Next: D.A. Davidson analyst Gil Luria said the talent drain from Google raises concerns that it’s losing the war for talent at the frontier of AI. Both Anthropic and OpenAI are planning IPOs, adding to the competition not just for talent but for investor dollars.

—Kit Norton

The hottest trade of 2026 was going sour on Tuesday as investors continued to ditch technology stocks. The selloff in South Korea’s flagship KOSPI could be a bad omen for red-hot Micron Technology shares—and a sign that a recent selloff is gathering pace.

• The KOSPI slumped 10% on Tuesday. The index is still up 95% this year despite the double-digit losses, having been powered higher by the artificial-intelligence boom.

• The KOSPI’s two largest members by total market capitalization are memory-chip makers Samsung and SK Hynix, and the boom in construction of data centers has driven up demand for their products. But both stocks fell more than 12% on Tuesday.

• There wasn’t one obvious factor driving Tuesday’s selloff. Investors are worrying that so-called hyperscalers won’t be able to continue spending at this rate on AI.

• The market is also fretting that the Federal Reserve will raise interest rates more than once in 2026 to combat inflation. Higher borrowing costs tend to hurt tech stocks because they make yield-bearing investments such as bonds more attractive and chip away at future cash flows.

What’s Next: The tech weakness in Asia seems set to extend into U.S. trading, and Micron looks likely to get caught up in the selloff. The memory-chip maker’s shares—up an eye-popping 234% so far this year—slumped about 7% ahead of the opening bell on Tuesday.

—Callum Keown and George Glover

Amazon’s Prime Day, the annual sales event that starts today and goes through Friday, will be the first for Alexa for Shopping, a new AI assistant to help people browse products on the website and app, and a litmus test for how much of a boost agentic shopping tools can boost transactions. 

• In a major update from Amazon’s previous generative-AI chatbot Rufus, Alexa for Shopping can put things in customers’ carts or track price changes on items they’re eyeing, shortening the time it takes for consumers to find what they need or locate relevant discounts, according to Adobe.

• Amazon launched the Alexa assistant in May, just as more people are using AI for their everyday shopping. A recent Adobe survey of 5,000 U.S. shoppers found that 39% of consumers have used AI for online shopping, and 85% of them said it improved their experience.

• Amazon is investing massive amounts of capital to build the physical infrastructure to power AI, as well as spending big on developing its own AI technology. Investors are hoping to see returns on these investments, including proof that shoppers are spending more on the Amazon website.

• BofA Securities analyst Justin Post is confident that Prime Day will show that Alexa for Shopping leads to increased consumer spending. He believes that Alexa for Shopping will be an “essential tool in protecting direct traffic for Amazon, as well as enabling higher conversion rates and driving incremental spend.”

What’s Next: Adobe’s latest forecast expects average daily online spending to be 84% higher during Prime Day than overall average daily spending in June. Post thinks the AI assistant can generate over $200 billion in incremental gross merchandise value by 2035, and $20 billion of incremental retail profit.

—Angela Palumbo and Janet H. Cho

Chevron made its biggest move yet in the AI data center sector, with a 20-year agreement to provide power to Microsoft’s facilities from its new natural gas plants in Texas. It means steady cash flows for Chevron for years, insulating it from the boom-and-bust nature of the oil patch.

• Chevron agreed to build 2.67 gigawatts of power capacity, enough to serve more than one million homes, from natural gas power plants in West Texas that will send electricity to Microsoft’s data centers at the same site.

• The power plants will operate fully disconnected from the grid, but may eventually be connected. Most of the turbines will be made by GE Vernova, supplemented by turbines made by a subsidiary of Caterpillar.

• Chevron, working with investment firm Engine No. 1, has been talking about the project in general terms for over a year. Chevron has enormous natural gas reserves in the region, but market prices are weak, because too few pipelines are available to transport out of the state.

• Natural gas in Texas has traded for negative prices in recent months, meaning producers must pay companies to take it away. Chevron has focused on signing longer-term deals for its gas, so it doesn’t have to sell into the spot market. 

What’s Next: Chevron and Microsoft haven’t technically reached a final investment decision, but the Microsoft transaction makes it all-but-certain it will happen, assuming it gets the necessary permits. Chevron estimates it can make annual returns in the midteens from the agreement.

—Avi Salzman and Janet H. Cho

The Trump administration is fast-tracking quantum computing, with President Donald Trump signing orders to develop the first-ever quantum computer for scientific research and accelerate the government’s migration to post-quantum cryptography by 2031.

• The computer for scientific research will be located in a national laboratory by 2028. Michael Kratsios, Trump’s top advisor on science and technology policy, says it’s a top priority for the administration alongside AI and nuclear technology.