📰 Key Takeaways

Alphabet reported solid revenue growth in Q3, with cloud sales hitting a record for the quarter and usage of Google’s AI services and in-house AI chips still climbing. But the stock plunged more than 10% at one point in after-hours trading once the results came out. The main reason is capex: Google put $51.4 billion into AI infrastructure in a single quarter and ended the quarter with negative free cash flow of $9 billion. The market read that as a warning sign, not as confidence in continued AI investment.

Samsung Electronics got a similarly cold reception. The South Korean chip giant said that, thanks to memory chip demand driven by the AI buildout, it became the first tech company in the world to post quarterly operating profit above 100 trillion won, yet its stock still fell after the results. The author points out that investors are increasingly worried about the massive AI spending at companies like Google, and they want real returns now, not five years from now. Chinese AI stocks have also dropped sharply in recent weeks on profitability concerns.

Another uncertainty is the US midterm elections: on November 3, about 40 local ballot measures related to data center development will go to voters across the US, and more voters than before are opposed to AI and data centers. The results will ripple through Asia’s tech supply chain, from Taiwanese server makers to South Korean chipmakers.

On top of that, PC makers like HP, ASUS, and Acer are moving some production lines from Thailand and Vietnam back to China to cope with soaring component costs, and are switching to a new generation of contract manufacturers. See the original article for details.

💬 JudyAI Lab Take

Alphabet and Samsung both turned in impressive results in the same quarter, and both stocks fell anyway. That tells us the market’s standard for judging AI investment is changing. Google’s cloud sales hit a record, but it spent $51.4 billion on AI infrastructure in one quarter, free cash flow was negative $9 billion, and the stock dropped more than 10% at one point after hours.

For AI builders, this case points to a trend: investors want real returns now, not five years from now. Samsung rode AI-buildout-driven memory demand to become the first tech company with quarterly operating profit above 100 trillion won, and its stock still fell after the results. Strong demand no longer automatically earns applause. On the other side, around 40 local ballot measures on data center development are going to voters in the US on November 3, and more voters than before oppose AI and data centers. Compute costs, cash flow pressure, and community acceptance are all tightening the room for AI infrastructure expansion at the same time. It’s a reminder that, from big companies to solo developers, the room to tell a story with just “spend more, use more compute” is shrinking, and proving unit costs and real output early will matter more and more.

My suggestion for this week: run the numbers on your own AI project. What does each call cost, and what output or savings does it bring in? Start by finding the piece with the biggest spend and the fuzziest return.

📅 Original Article Info

🔗 Further Reading