When I watched Eric Schmidt’s TED Talk on the AI revolution, I had one of those rare “aha” moments—the moment when pieces of the puzzle suddenly fall into place and a clearer picture emerges. The former Google CEO and chairman wasn’t just talking about technology; he was painting a vision of our collective future that is as fascinating as it is challenging.
What particularly impressed me was Schmidt’s assertion that the AI revolution is not overhyped, as many believe, but actually underestimated. At a time when AI headlines dominate the media, this sounds paradoxical at first. But his arguments are convincing.
Mit dem Laden des Videos akzeptieren Sie die Datenschutzerklärung von YouTube.
Mehr erfahren
The quiet moment that changed everything
Schmidt begins his narrative with a historic moment that took place in 2016 but went largely unnoticed by the general public. The AI system AlphaGo made a move in the game of Go that no human had ever thought of in the game’s 2,500-year history. An algorithm had created something genuinely new.
“How is it possible that our computers could develop something that humans never thought of? I mean, this is a game played by billions of people,” Schmidt asks rhetorically. For him, this moment marked the real beginning of the AI revolution – long before ChatGPT electrified the public.
What fascinates me about this story is that it’s not about AI imitating or surpassing humans, but about AI thinking in fundamentally different ways and finding solutions that lie beyond our human horizon.
Why we underestimate the AI revolution
Most of us probably had our first “wow” moment with AI through ChatGPT or similar language models. However, Schmidt argues that we are only scratching the surface when we view AI primarily as an impressive writing tool.
“Since ChatGPT, advances in reinforcement learning, which helped develop AlphaGo, have enabled us to carry out planning,” Schmidt explains. He describes how modern systems such as OpenAI’s ‘3o’ or DeepSeek’s ‘R1’ can perform complex planning processes – moving forward and backward, weighing options, developing strategies.
As I thought about it, it became clear to me that we are not facing incremental improvements to existing technologies, but a fundamental change in the way machines solve problems. Schmidt himself uses this technology to familiarize himself with complex subject areas – such as the rocket technology of a company he has acquired. According to him, the computing power contained in just 15 minutes of AI use is “extraordinary.”
From language to strategy: the next stage of evolution
What comes next? Schmidt outlines a fascinating development: ”You see the transition from language to language, then from language to sequence, as is done in biology. Now you’re essentially doing planning and strategy.”
The vision of the future he paints is a world in which AI agents control entire business processes, communicate with each other, and coordinate complex tasks. These systems will not only react, but also proactively plan, discover, and collaborate.
I thought about all the repetitive business processes that are still performed by humans today – and what a world might look like in which these tasks are taken over by intelligent systems, while humans can focus on more creative and fulfilling activities.
The hurdles along the way: energy, data, and knowledge
But this vision faces tangible challenges. Schmidt identifies three main problems:
- Energy requirements: “We need another 90 gigawatts of electricity in America,” explains Schmidt – that’s equivalent to 90 nuclear power plants. The computing power required for advanced AI systems is enormous and growing exponentially.
- Data scarcity: “We’ve used up the data, so we have to start generating it.” The publicly available data on the internet is no longer sufficient to train the next generation of AI models.
- Limits of knowledge: Perhaps the most fascinating challenge concerns the ability to discover something truly new. Schmidt explains how scientific breakthroughs often come about: A brilliant person recognizes a pattern in one area that has similarities to a completely different area and transfers tools from one to the other. “Today, our systems cannot do that.”
These hurdles make it clear that the AI revolution is indeed a marathon—with technical, logistical, and conceptual challenges that need to be solved.
The geopolitical race: US vs. China
One aspect that gave me particular pause was Schmidt’s analysis of the geopolitical dimension. He sees a “defining battle” between the US and China for AI supremacy that could have far-reaching consequences.
While the US is focusing on closed, secure models, China is tending toward open-source frameworks such as DeepSeek – efficient systems that are accessible to everyone. “China’s open-source approach could democratize AI… or use it as a weapon,” warns Schmidt.
What is at stake is nothing less than “mutually assured disruption.” If one nation gains a decisive lead in the development of super-intelligent AI, rivals could resort to sabotage. “Imagine hacking or even bombing data centers,” Schmidt says gravely.
These thoughts reminded me of the tensions of the Cold War – except that this time we have no diplomatic framework to manage AI-driven conflicts. “We are replaying 1914,” Schmidt warns, referring to Kissinger’s fear of accidental war. “We need rules before it’s too late.”
Ethical dilemmas: security vs. surveillance
The dual-use nature of AI – useful and potentially dangerous at the same time – forces us to make difficult decisions. Preventing misuse (e.g., for biological weapons or cyberattacks) carries the risk of creating a surveillance state.
Schmidt advocates cryptographic “proofs of humanity” without compromising privacy: “Zero-knowledge proofs can verify humanity without revealing identities.”
He also emphasizes the importance of ‘meaningful human control,’ referring to the US military doctrine. At the same time, he criticizes overly strict regulation: “Stopping the development of AI in a competitive global market is naive. Instead, we should build guardrails.”
Finding this balance—between innovation and security, between progress and control—seems to me to be one of the greatest social challenges of our time.
The brightest promises: curing diseases, deciphering physics, creating billions
Despite all the risks, Schmidt is optimistic. AI could eradicate disease by accelerating drug development: “A nonprofit organization wants to map all human ‘drug’ targets in two years.” Another startup claims to reduce the cost of clinical trials tenfold.
In education, AI tutors could personalize learning for every child, in every language. Schmidt asks rhetorically, “Why don’t we have every person on the planet with their own tutor in their own language to help them learn something new?”
These visions of a better future remind us that the AI revolution brings not only challenges, but also enormous opportunities.
A marathon, not a sprint: How we should deal with change
For those who feel overwhelmed by the rapid changes in technology, Schmidt offers some advice: Treat this transition as a marathon, not a sprint.
“Remember that this is a marathon, not a sprint,” says Schmidt. ”My advice to all of you is: Ride the wave, but ride it every day. Don’t think of it as episodic.”
He emphasizes that each of us has a reason to use this technology—whether as an artist, teacher, doctor, businessperson, or technician. “If you don’t use this technology, you will not be relevant compared to your peer groups and your competitors and the people who want to be successful. Embrace it, and embrace it quickly.”
These words struck me deeply. They reminded me that technological change does not require a one-time adjustment, but rather a continuous willingness to learn and evolve.
Conclusion: The most significant development in 500-1000 years
Schmidt concludes with a remarkable assessment: “The arrival of this intelligence, both at the AI level, AGI, or general intelligence, and superintelligence, is the most important thing that will happen in human society in about 500 years, maybe 1,000 years, and it’s happening in our lifetime.”
After watching this fascinating conversation, I am convinced that the AI revolution is indeed a marathon, not a sprint. It requires perseverance, strategic thinking, and a willingness to continuously adapt. It brings enormous challenges—technical, ethical, geopolitical—but also unprecedented opportunities.
What impressed me most is Schmidt’s ability to recognize both the transformative power of this technology and to analyze its risks soberly. His message is clear: we are at the dawn of a new era, and how we shape it will define the next centuries. It’s a marathon, and we’ve only just begun.
Source: https://www.youtube.com/watch?v=id4YRO7G0wE
