Former Google CEO Eric Schmidt dives into the transformative power of artificial intelligence, from superintelligent agents to global power shifts, in a riveting discussion that echoes historical turning points.
In a packed auditorium at the RAISE Summit in France, Eric Schmidt, the tech visionary who once steered Google through its explosive growth, took the stage to deliver a message that felt both exhilarating and eerie. Drawing from his co-authored book The Age of AI: And Our Human Future—penned alongside the late diplomatic giant Henry Kissinger and Microsoft researcher Craig Mundy—Schmidt painted a picture of artificial intelligence not as a mere tool, but as the harbinger of a new epoch in human history. It’s a bold assertion, one that harkens back to the Enlightenment, when humanity shifted from blind faith to reasoned inquiry. But as Schmidt sees it, we’re on the cusp of something even more profound: a nonhuman intelligence that could outthink us all. What if, in our rush to innovate, we’re unleashing forces we can’t fully control? It’s a question that lingered in the air long after his talk ended.
Schmidt’s appearance wasn’t just a keynote; it was a wake-up call. France, with its burgeoning tech scene, provided the perfect backdrop. For years, Schmidt has collaborated with French innovators, and he beamed with pride at the country’s vibrant ecosystem. Yet, his words carried a sense of urgency. Leaders, he argued, are still misunderstanding the scale of this shift. They see AI as flashy demos like ChatGPT, a breakthrough from two years ago that’s already ancient history in tech terms. The real game-changer? Systems that reason, remember, and act autonomously—agents that could redefine everything from building homes to running governments.
The Dawn of a New Epoch: Echoes of the Enlightenment
To grasp Schmidt’s vision, we need to step back into history. The Enlightenment, that 18th-century intellectual revolution sparked in salons across Europe—think Voltaire in France or Locke in England—marked humanity’s pivot from divine authority to rational thought. We began questioning, experimenting, and building societies based on logic rather than lore. Kissinger, Schmidt’s dear friend and co-author who passed away in 2023 at age 100, drew parallels here. In their book, they posit that AI represents a similar rupture: the arrival of a superior reasoning entity.
But why call it an “epoch”? Schmidt is convinced we’re barreling toward Artificial General Intelligence (AGI), where machines don’t just mimic tasks but exhibit free will-like behavior—deciding, learning, and pursuing goals independently. Beyond that looms superintelligence, smarter than all humans combined. Imagine a system proving theorems we know are true but can’t comprehend ourselves. Is that genius or sorcery? Henry would have called it the latter, warning it might spark fear-driven backlash.
Schmidt’s concern isn’t abstract. He consults on national security and sees AI reshaping warfare, economies, and democracies. Will our systems of governance hold up when machines outpace human decision-making? It’s a rhetorical flourish, but one that hits home: in an era of deepfakes and algorithmic biases, could AGI erode trust in elections or amplify authoritarian control?
The San Francisco Consensus: A Bold Bet on the Future
Shifting gears, Schmidt introduced a term that’s buzzing in tech circles: the “San Francisco Consensus.” It’s not some formal accord, but a shared belief among Bay Area elites—think startup founders and AI pioneers—that the world will transform in two to four years, with three as the sweet spot. Schmidt chuckles at the hubris, noting a consensus isn’t always correct, but the logic is compelling.
It starts with large language models (LLMs) like ChatGPT, which handle conversation brilliantly. Add memory and reasoning, and you get “agentic” AI—systems that chain actions together. Schmidt’s example? Building a house: one agent scouts land, another navigates regulations, a third designs, and yet another handles the inevitable lawsuits. It’s humorous, but spot-on. This isn’t sci-fi; it’s workflow automation on steroids, poised to disrupt every industry.
He spotlighted recent advances, like OpenAI’s o1 model (which he called “03” in a slip, but the point stands), that reasons step-by-step, forward and backward, tackling problems with mind-blowing precision. Google’s math AI now rivals top grad students, hitting 90th percentile in fields like physics. These aren’t incremental tweaks; they’re leaps toward recursive self-improvement, where AI learns from itself at combinatorial speeds humans can’t fathom.
Schmidt’s timeline? He pegs superintelligence at six years, give or take, while the consensus says three. Either way, it’s alarmingly soon. And governance? We’re scrambling. Math suggests we might “contain” superintelligence, but Schmidt admits uncertainty. By year’s end, he predicts, we’ll interact with self-learning systems daily. Exciting? Absolutely. Worrisome? You bet—especially when these models guzzle thousands of times more power than a simple search.
From AGI to Superintelligence: Navigating the Unknown
Diving deeper, Schmidt clarified the jargon that’s often tossed around like confetti. AGI is general smarts: waking up with “free will,” charting your own path. It’s not here yet, but Schmidt forecasts four to six years. Superintelligence follows, outstripping collective human intellect. The litmus test? Proving truths beyond our grasp.
This isn’t just philosophical musing. Schmidt’s national security lens reveals high-stakes risks. In a world of network effects—where winners like Google or Meta snowball advantages—AI could create unbridgeable gaps. Picture two nations racing: one deploys a million AI researchers (no breaks, no pizza), innovating exponentially. The loser might preemptively strike, fearing permanent inferiority. It’s a “race condition” that could ignite conflicts.
Geopolitically, this echoes Cold War tensions but with silicon instead of nukes. Schmidt, who chaired the U.S. National Security Commission on AI, warns of democracies’ fragility. Can they survive AGI’s manipulations? Personalized propaganda or autonomous weapons might tip the scales. And on a personal note, as someone who’s watched tech evolve, I can’t help but reflect: we’ve chased progress for decades, but at what cost to our humanity?
The Compute Boom: Is It a Bubble or a Revolution?
No AI talk skips the money. Billions pour into data centers—OpenAI’s $100 million raise pales against Meta’s $20 billion compute spend. Schmidt quips it’s the new “Capex moat,” but is compute king over talent?
Industry execs whisper of overbuilding, predicting overcapacity in years. Yet each claims they’ll thrive while rivals flop—a classic bubble signal. Schmidt pushes back: if reasoning chains and inference explode, we’re underinvested. These models demand insane resources, but history shows software always fills hardware voids. Remember the old quip about Intel’s chips speeding up only for Microsoft to bloat them with features? Same here.
Crediting OpenAI’s Sam Altman, Dario Amodei (now at Anthropic), and Mira Murati for kickstarting foundation models, Schmidt calls it a “historically important discovery.” It’s birthed an industrial shift, powering scale-free fields like math (endless conjectures) and software (code on demand). Soon, we’ll dictate programs, rendering traditional coding obsolete. Schmidt, a language design PhD, laments this with wry humor—his life’s work upended in a lifetime.
But positives abound: AI could crack climate change, medicine, biology. On the flip side, cyber threats multiply as systems generate attacks relentlessly.
Geopolitical Shifts: The Chip Race and China’s Open-Source Gambit
Schmidt’s mantra: “Silicon is strategy.” The U.S.-China AI race is intensifying, with export controls on chips like Nvidia’s creating chokepoints. Yet China adapts, pumping out open-source models via outfits like DeepSeek, likely government-backed.
Contrast: U.S. firms hoard closed-source tech in mega data centers; China floods the world with open weights. Result? Open source dominates in underserved nations—over 100 countries we overlook. Schmidt flags the irony: the West leads, but Chinese AI might power the global majority.
Europe shines too, with more open-source math contributions than the U.S., per Schmidt’s chats at the summit. It’s a multipolar world, where alliances matter. Geopolitically, this could realign power, much like the Space Race did, but with intangible code as the prize.
Lessons from the Mobile Era: Speed or Be Sidelined
Reflecting on Google’s mobile pivot—Android now claims 90% market share—Schmidt admits timing errors. They didn’t push hard enough, missing how GPS birthed Uber or phone numbers became IDs for WhatsApp.
His advice? Act now, fast. AI’s crowded field demands relentless product focus. Hesitate, and you’re bypassed. Startups like Cursor, hitting $500 million ARR in months (or was it $100 million? The excitement blurs details), exemplify this velocity.
Conclusion: Embracing the Inevitable with Eyes Wide Open
As Schmidt wrapped, joined briefly by founder Andrew Feldman, the crowd buzzed. AI isn’t just tech; it’s a mirror to our ambitions and fears. We’re engineering a new intelligence that could elevate humanity—or eclipse it. Schmidt’s blend of optimism and caution feels right: celebrate the vibrancy, like France’s tech rise, but prepare for the unknowns.
In the end, will we harness this epoch for good? Or let it unravel us? History’s great shifts—the Enlightenment, the Industrial Revolution—brought progress amid pain. AI might be our biggest yet. As Schmidt urges, let’s get organized—for the good and the bad. After all, the future isn’t waiting.
