What Is The San Francisco Consensus, Silicon Valley's AI Prophecy?

Forbes

23-07-2025

Eric Schmidt, billionaire and co-founder of Schmidt Futures, at the Raise summit in Paris, France, ... More on Tuesday, July 8, 2025. The annual conference gathers global leaders and key speakers in tech and AI. Photographer: Nathan Laine/Bloomberg © 2025 Bloomberg Finance LP
Just as the Washington Consensus dictated economic orthodoxy for developing nations throughout the 1990s, a new doctrinal framework is crystallizing among the power brokers of Silicon Valley. They have coalesced around a staggering prediction: Artificial intelligence will fundamentally transform all aspects of human activity within a mere three to six years.
This prophecy—dubbed the "San Francisco Consensus" by former Google CEO Eric Schmidt—transcends typical Silicon Valley hyperbole. It represents Silicon Valley's equivalent of Washington's economic rulebook, a shared conviction driving trillion-dollar investments and strategic priorities among those actively constructing our AI future.
Unlike the Washington consensus, which mandated fiscal austerity and economic restraint, Silicon Valley's calls for unprecedented investment and acceleration. Tech leaders are not advocating belt-tightening. They are pouring billions into AI infrastructure at a breathtaking pace, with the conviction that they are financing humanity's next evolutionary leap.
"I call it a consensus because it's true that we agree... but it's not necessarily true that the consensus is true," Schmidt said at the RAISE summit in Paris earlier this month. This paradox captures both the fervor of belief and the inherent uncertainty in technological prophecy.
What makes this consensus consequential is not the belief itself, but who believes it. They are the architects constructing our digital future. When the people controlling trillions in capital and computing resources share a conviction, reality has a curious way of conforming to their expectations.
Schmidt frames the San Francisco Consensus as three simultaneous technological revolutions amplifying each other: The Language Revolution Large language models like ChatGPT have inverted the traditional human-computer relationship. Instead of humans learning machine languages, machines now understand ours. Language has become the universal interface for human-AI interaction.
This goes beyond AI mimicking human prose, although watching it compose Shakespearean sonnets about cryptocurrency is entertaining. It represents a fundamental shift in accessibility, democratizing technological power in ways previously unimaginable. The priesthood of programmers is giving way to a world where anyone who can articulate their needs can harness computational power. The Agentic Revolution If the Language Revolution concerns AI understanding us, the Agentic Revolution involves acting independently in our world. We are witnessing the transition from AI as glorified calculators to autonomous actors. They were tools. Now they are agents. Schmidt envisions a near future where entire business workflows—from purchasing real estate to resolving contractual disputes—are managed by interconnected AI systems. Imagine digital agents negotiating, collaborating, and executing complex tasks with the efficiency of a bureaucracy but without the delays or regulations.
This shift from passive tools to active agents represents a profound reconceptualization of how computational systems integrate into human affairs. The question is no longer what computers can calculate, but what they can achieve. The Reasoning Revolution
The centerpiece of this technological triptych is AI's metamorphosis from pattern-matcher to genuine reasoner. New models by Google as well as OpenAI outperform graduate students in mathematics. This development might alarm academic departments but delight undergraduates struggling with differential equations.
This represents the leap from machines that can mimic human outputs to systems that genuinely cogitate—plotting, analyzing, and problem-solving with remarkable acuity. We are witnessing the transition from impressive mimicry to unsettlingly human-like thought.
The consensus timeline is breathtakingly compressed: Within months: Programming jobs and skilled roles begin vanishing faster than Taylor Swift concert tickets, transforming labor markets overnight. Within 3-6 years: Widespread artificial general intelligence (AGI) emerges, with machines rivaling our brightest minds across intellectual domains. In a decade: Artificial superintelligence (ASI) arrives. AI systems make discoveries beyond human comprehension. Picture explaining quantum mechanics to your golden retriever. Now imagine you are the retriever! Reality Reckons
However, not everyone is drinking the algorithmic Kool-Aid. Why? Silicon Valley's prophecy graveyard is crowded. Flying cars, ubiquitous VR, Web3. The list goes on. Detractors say the San Fran Consensus could be another case of technological exuberance outrunning reality. They have a point here.
After all, the epistemic bubble is real. When surrounded by true believers in Palo Alto, reality distortion fields form. This consensus may reflect groupthink among tech elites accustomed to deference.
So are the hurdles. Reliability issues make current AI resemble temperamental toddlers at times. Replit's CEO, Amjad Masad, publicly apologized after its autonomous agent erased a production database. ChatGPT is said to have caused multiple hospitalizations for Jacob Irwin, a 30-year-old on the autism spectrum, because it fueled his time-travel fantasy. Are these bugs to fix? Maybe. But these instances represent fundamental challenges to trustworthy AI that the Consensus promises. The Consequential Consensus
Whether prophetic or destined for technology's unfulfilled promises museum, the San Francisco Consensus is already redirecting capital flows and strategic priorities. Look at the infrastructure gold rush. Specialized chips and data centers are proliferating at breakneck speed. The self-fulfilling power of the collective belief of Silicon Valley elites is at full display here. Perhaps we will read about the San Francisco Consensus in textbooks someday.