The AI boom isn't a bubble — it's barely begun, Goldman Sachs says (www.businessinsider.com)

Goldman Sachs says the AI boom is just beginning and could create trillions in economic value, arguing current spending is small relative to the upside. In a research note the bank estimates generative AI could add roughly $20 trillion to the U.S. economy over time, with about $8 trillion accruing to companies, and a baseline scenario that full adoption would lift economy‑wide labor productivity by ~15% over a decade. Despite record investment in chips, servers and data centers, Goldman judges AI capex is still under 1% of GDP (about $300 billion annually in 2025), versus the 2–5% GDP peaks seen in prior revolutions like railroads, electrification and the dot‑com era. Technically and strategically, Goldman highlights two drivers: measurable productivity gains from deployed AI and the massive compute required to unlock further value, which support continued investment even as hardware costs decline. But the note warns that early spenders aren’t guaranteed long‑term winners—rapid hardware depreciation, fast model improvement, multi‑model adoption by customers, and historical patterns (where later entrants captured better returns after overbuilds) could reshape market leadership. The takeaway for AI/ML practitioners and investors: meaningful upside remains, but timing, complementary assets, and capital allocation will determine who ultimately captures the value.