Tech community converges on 2035 as AGI consensus date — but is it just a convenient decade?

Summary

The tech community has coalesced around 2035 as the consensus date for AGI, but the alignment may say more about psychology and incentive structures than confidence in the timeline itself.

The 2035 frame is suspiciously convenient. It matches a ten-year venture fund lifecycle—long enough to feel serious, short enough to avoid radical life changes today. If you believe AGI arrives in three years, you're probably not writing blogs about it; you're starting a macro hedge fund or working at a lab to influence outcomes. If you don't believe it's coming at all, you're not writing about AGI either. The consensus around a decade out sits in a sweet spot where people can discuss the scenario without committing their actual behavior to it.

Ten years is too distant for most people to plan around. Someone in their twenties might want to own a home by 2035, but that's a want they have today anyway. Actually restructuring your life around an uncertain AGI scenario is very difficult. The dialogue has shifted from "golden retriever maxing"—the 2024 joke about maximal leisure in your final pre-AGI years—to a more diffuse focus on "escape," whatever timeframe individuals pick. But the escape narrative persists: one year, five years, or now, a decade. The date changed; the anxiety didn't.

Revealed preference tells the story. Despite holding bullish AI views, people typically spend their time podcasting rather than starting biotech companies or working at frontier labs. The gap between stated beliefs and actual allocation of time and capital is the tell.

Vibe coding with AI is legitimate. The criticism that it lacks rigor misses the point. Vibe coding is closer to playing Minecraft or building with Legos than to writing production software. The alternative critics propose—pen-and-paper coding, pure math—doesn't track with how most kids actually learn engineering. Vibe coding tools would have transformed iPhone app experiments in middle school, compressing what took months into faster iteration.

The broader push for entrepreneurial literacy matters. Most kids are offered a narrow menu of career archetypes and "businessman" remains vague. Entrepreneurship as a frame for how to think about your life is hidden unless someone surfaces it. The call to vibe code is less about technical preparation and more about access to a mental model that isn't usually served up. Tim Ferriss's The 4-Hour Workweek worked for the same reason: it wasn't a technical manual, it was permission to see business-building as a viable life option.

SoftBank and the OpenAI accounting signal

SoftBank booked $8 billion in gains on OpenAI shares in the last quarter, then sold its entire $5.8 billion NVIDIA stake to fund the investment it's already marked up. The move is not illegal but is nontraditional. SoftBank is recognizing a gain on an investment it hasn't fully paid for. Vision Fund 2 committed capital but hasn't deployed it all. This is a pattern in VC, but the circularity is visible: profit on OpenAI, sell your AI chip exposure, buy more OpenAI.

Masayoshi Son has moved this way before. He bought a 5% NVIDIA stake through Vision Fund in 2017, exited completely in January 2019 for $3.6 billion, and would have $200 billion today. The current sale suggests impatience again or a genuine conviction that OpenAI upside outweighs continued NVIDIA exposure.