Evidence from 2024 and early 2025 points against Chamath’s prediction that bootstrapped/profitable startups would be the main business winners of 2024 because AI/compute would become so cheap that small teams could rapidly clone incumbents.

Key points:

• The biggest financial winners of the 2024 AI boom were large, capital‑intensive, often unprofitable, VC‑ or public‑market–funded companies, not bootstrapped startups. U.S. venture funding rebounded to about $209B in 2024, with AI startups capturing a record ~46% of that and raising giant rounds (e.g., OpenAI, xAI), many still unprofitable and heavily compute‑dependent. (reuters.com) Market commentary on “AI winners” for 2024–25 consistently highlights Nvidia and other large public players, not bootstrapped firms, as the primary beneficiaries of the AI wave. (investing.com)

• Frontier AI compute did not become “cheap” in 2024; if anything, costs at the cutting edge rose sharply. A 2024 analysis estimates the amortized cost of training frontier models (e.g., GPT‑4‑class systems) has been growing at ~2.4× per year since 2016, with single training runs costing tens of millions of dollars and projected to exceed $1B by 2027—only accessible to the very well‑funded. (arxiv.org) In 2024, Nvidia’s H100 GPUs were selling in the ~$25,000–$30,000 range (and more on secondary markets), underscoring that top‑tier AI compute remained extremely expensive, favoring deep‑pocketed incumbents and VC‑backed firms rather than small bootstrappers. (en.wikipedia.org)

• While bootstrapped and profitable startups did see notable success, they were exceptions, not the dominant “biggest winners.” Articles highlight standout bootstrapped AI companies such as Surge AI (reported as “well north of $1B” in 2024 revenue) and Midjourney (hundreds of millions in revenue, profitable, and fully bootstrapped), as well as profitable bootstrapped firms in energy and design/build. (aimmediahouse.com) These prove the viability and appeal of bootstrapping in the AI era—but they are framed as remarkable outliers amid an ecosystem still dominated, in dollar and market‑share terms, by heavily funded AI ventures.

• The funding environment did push more founders toward bootstrapping and profitability, but even bullish analyses don’t claim bootstrappers clearly outperformed VC‑backed peers in 2024. PitchBook‑linked commentary in early 2024 predicted that a brutal funding environment "may well" make 2024 a year of the bootstrapped founder, largely because VC supply tightened outside of hot AI segments. (biohealthinnovation.org) Later data from lenders and CB Insights suggest bootstrapped startups often grow efficiently and reach profitability faster than VC‑backed ones, but these are structural, multi‑year observations—not specific evidence that bootstrappers were the biggest winners of 2024 in terms of market impact or aggregate value creation. (sidetool.co)

• Rapid disruption in “months instead of decades” did occur in some niches, but mostly via well‑funded AI players rather than small bootstrapped clones. For instance, Anysphere’s Cursor AI coding assistant saw explosive jumps in valuation and revenue, but as a VC‑backed startup, not a bootstrapped one. (theinformation.com) The broader AI application wave did enable small teams to ship products quickly, but the largest disruptive forces and market share shifts remained concentrated in big‑tech and VC‑heavy firms.

Taken together, 2024 did validate some of the spirit of Chamath’s view (that lean, profitable and even bootstrapped AI companies can succeed and that AI tools compress development timelines). But the strong form of the prediction—that bootstrapped/profitable startups would be the biggest business winners of 2024 because AI/compute became so cheap that they could rapidly clone incumbents—is not supported by the actual funding flows, cost structure of frontier AI, or lists of top corporate and market winners. Hence, the prediction is best judged wrong.