It was a quiet day from Serenity, the stock-picker whose reconstructed positioning this site tracks. He published no new trades and named no tickers, instead offering a brief macro reflection that ties back to his broader worldview.

The trigger was a widely circulated comment from Elon Musk suggesting artificial intelligence could surpass human capability across essentially all fields by the end of next year, or by 2028 at the latest. Serenity connected that prediction to a framework he has referenced before: Anthropic's paper on the economic implications of advanced AI, which he reads as sketching out 2026–2028 as the heavy infrastructure buildout phase, followed by 2028–2030 as a period of unusually rapid economic growth.

His takeaway was a timing one. On that map, he argued, we are still only in the early second half of 2026 — meaning, in his view, the buildout is nowhere near finished and the trade still has room to run.

That reading is the connective tissue for almost everything Serenity does. His investing approach is built on "buying what's upstream of NVIDIA": he treats the chip giant's roadmap — high-bandwidth memory, advanced packaging, co-packaged optics (putting the optical engine right next to the switch or GPU die to save power and lift bandwidth) — as a signal of where demand will eventually land, then hunts for the small, obscure suppliers that the whole chain quietly depends on. He calls these "chokepoints": critical upstream vendors with few substitutes.

His dominant theme is AI photonics — the lasers and specialty substrates, often built on indium phosphide, that feed the optical interconnect roadmap. Around that sit his other sleeves: a memory supercycle bet on constrained DRAM and NAND flash pricing, advanced-packaging tooling, the electrical-grid bottleneck that data-center power demand creates, AI cloud capacity names, and a defense-and-drones basket. All of them share the same underlying premise — that the AI infrastructure spend is large, sustained, and still building.

Seen in that light, today's post reads less as a new idea and more as a conviction check. By anchoring to a "still early" position on the buildout timeline, Serenity is signaling that he sees no reason yet to fade the multi-year theme underpinning his portfolio. It is worth noting the caveats such a stance carries: forecasts from public figures about AI's pace are speculative, and a thesis that leans on the buildout continuing for years is exposed to any slowdown in that spending. Serenity offered no timeline adjustments, new entries, or exits today.

With no fresh positions to report, the day amounts to a restatement of philosophy rather than action — a reminder of the macro clock he believes he is trading against, and his contention that the hands have not yet moved as far as the headlines might suggest.

*This is derived commentary reconstructed from public posts, not investment advice, and not the subject's own words.*