人工智能吞噬世界

Benedict EvansNovember 2025www.ben-evans.com The next platform shift Every 10-15 years, a platform shift reshapes technology What happens in a platform shift? Who is affected, and how much? Dominance is won and lost Microsoft dominated the PC era, but when the centre of gravity shifted to smartphones it became irrelevant Early leaders often disappear The ‘first’ was not the winner in PCs, browsers, search, social or smartphones How will the new thing work? We don’t know For every new platform, we forget how many ideas failed and how unclear everything was Noise, hype, anti-hype When things are exciting, people get excited This often brings bubbles People draw straight lines on log scale charts They forget that exponential growth isgenerally an exponential curve And always say “this is different” The trouble is, they’re generallyright- everybubble is different! But it can still be a bubble But when the dust settles, the world has changed The internet has gone from the New Thing to a basic part of daily life US heterosexual couples who met online, by year of meeting New platforms mean new tools (and new revenue) SaaS means the typical large enterprise in the USA now uses 4-500 apps One way this platform shiftisdifferent, though “The race to AGI is afoot”Sergey Brin For PCs, the web or smartphones, we knew thephysical limits of what could happen next year “AGI needs multiplefurther breakthroughs”Demis Hassabis With LLMs, we don’t know how much betterthis could get Another platform shift, or more? We know this will get better, but we don’t know how much So how will the new thing turn out? If this is ‘only’ as big as mobile or the internet, that seems like enough Inside tech “The very worst case wouldbe that we have just pre-built for a couple of years” “The risk of under-investingis significantly greater thanthe risk of over-investing” Mark Zuckerberg, Q3 2025 Sundar Pichai, Q2 2024 Three years of FOMO in Big Tech “This is a huge new market, a huge threat to all our existing businesses, and we can’t miss it” FOMO drives a capex surge ~$400bn in 2025 for the big four alone (for comparison, global telecoms is ~$300bn) Planned 2025 growth almost doubled - in 2025 ~$350bn in 2025 for the big four alone (for comparison, global telecoms is ~$300bn) A new investment cycle US data centre construction overtaking offices OfficesRetailWarehouses *Data centres** Nvidia can’t keep up Trying to build a new Sun Microsystems (though China and hyperscalers’ own chips are coming up behind) Nvidia can’t keep up (neither can TSMC) TSMC unwilling/unable to expand capacity fast enough to meet Nvidia’s book US power backlogs becoming a major issue US power demand growth is ~2%, and AI might add 1% that’s hard to build fast (this is not an issue in China) “It’s been almost impossible to buildcapacity fast enough since ChatGPTlaunched” Kevin Scott, Microsoft CTO “We now expect the FY26 growth rate to behigher than FY25” - Microsoft “Capex dollar growth will be notably larger in2026” - Meta “We expect a significant increase in 2026” -Alphabet Data centre capacity triples? For $3tr? $5tr? More? Some very large numbers (although some of the ‘bragawatts’ may be more performative than real) “Three trillion dollars!” Annualised AI capex aspirations are a similar magnitude to mature global capital-intensive industries The hyperscalers can afford it… Big Tech cashflow has surged since the Pandemic, and most of that growth is going on AI capex The hyperscalers can afford it… up to a point Capital leases are not new, but they’ve got a lot bigger Up to a point Hyperscalers add leases and debt, while some analysts suggest Oracle’s cloud capex might be >100% of revenue OpenAI joins the club Announced commitments for 30GW+ ofcapacity at $1.4tr Aspiration for 1GW/week of new constructionat $20bn/GW = ~$1tr annually… Equivalent to 2/3 of total current global base,every year “Circular revenue” Without its own cashflows, OpenAI partnerswith Nvidia, Oracle, Softbank, petrodollars… OpenAI is buying Nvidia chips with Nvidia’scashflow… Which comes from the hyperscalers… and using Nvidia’s cash to turn AMD into anNvidia competitor, and pay Broadcom to makeits own chips… Rational actors? What would you do if your company was sitting on a bubble? Nvidia has $72bn of TTMFCF* and TSMC can’t keepup with demand Use your excess cash tobuy demand, FOMO andplatform lock-in OpenAI has mindshareand expensive stock, butno platform, infra or moat Swap your paper for hardassets and market position Oracle is a cash-generativelegacy business losing shareto cloud and now to AI Gear up and burn yourway into the new thing? Yes, but where has all this money got us? After three years, lots more science and engineering, but no real clarity on the shape of the market Far more models Every week - new models, new (problematic, gamed, saturated) benchmarks, new acronyms Models converge and leaders