Whenever I study Taiwan Semiconductor Manufacturing Company, I feel as though I am examining the industrial equivalent of the person in a group project who quietly completes everyone else’s work while the louder participants argue over the presentation. TSMC does not design the world’s most famous processors. It does not sell smartphones, build artificial-intelligence models, or manufacture graphics cards under its own consumer brand. Instead, it manufactures the advanced chips designed by many of the companies receiving all the attention. Apple dreams up a new processor. Nvidia designs another computational beast capable of making data centers glow like small artificial suns. AMD prepares its next attack on the server market. Qualcomm develops another mobile platform. Broadcom expands its custom silicon business. Then everyone eventually arrives at TSMC’s door carrying blueprints and extremely large purchase orders. That arrangement has made TSMC the dominant pure-play semiconductor f...
Alphabet’s search business is still growing while artificial intelligence rewrites how people find information. The real investment question is not whether Google survives. It is what survival costs—and whether the new version of Search can remain as profitable as the old one. For most of my adult life, “Google it” has been less a suggestion than a reflex. I do not announce that I am about to use a search engine. I simply open a browser, type half a thought into a box, and expect the accumulated knowledge of civilization to arrange itself helpfully before I lose interest. Google became so embedded in daily behavior that its brand stopped functioning like a company name and started behaving like a verb, a utility, and occasionally a substitute for consulting a qualified physician. Then generative artificial intelligence arrived and asked an impolite question: What if people no longer want ten blue links? What if they want one direct answer? What if they ask ChatGPT, Perplexity, Claude, ...