Ex-Google CEO: What Artificial Superintelligence Will Actually Look Like w/ Eric Schmidt & Dave B

The Imminence of Digital Superintelligence

Schmidt opens by emphasizing that digital superintelligence is not a distant fantasy but a near-term reality, likely to emerge within the next ten years. He highlights the concept of AI’s ability to generate its own scaffolding—essentially, the frameworks and structures that enable it to learn and improve autonomously—as a breakthrough expected around 2025. This recursive self-improvement, while not yet fully realized, marks a critical step toward AI systems that can independently advance their own intelligence and capabilities.

Schmidt describes the future AI polymath as a pocket-sized entity combining the intellectual prowess of historical geniuses like Einstein and Leonardo da Vinci. This polymath will be accessible to individuals, fundamentally transforming how knowledge and creativity are harnessed. However, he cautions that alongside these positive developments, there exists a negative domain involving risks such as biological and cyberattacks that could be enabled by superintelligent systems, underscoring the need for careful planning and regulation.

Energy and Infrastructure Challenges

A significant portion of the discussion focuses on the enormous energy requirements of AI development and deployment. Schmidt points out that the natural limit to AI’s growth is not chip technology but electricity. The current AI revolution demands vast amounts of power, with estimates suggesting the need for an additional 92 gigawatts in the U.S. alone—equivalent to the output of multiple nuclear power plants. Yet, the construction of new nuclear facilities is slow and unlikely to meet this demand in time.

While fusion and small modular reactors (SMRs) offer hope for future energy solutions, their timelines do not align with the immediate needs of AI infrastructure. Schmidt notes that companies like Meta, Google, Microsoft, and Amazon are securing long-term nuclear energy contracts to power their data centers, effectively taking on roles traditionally held by utilities. The energy race is also geopolitical, with China’s abundant electricity resources positioning it as a formidable competitor in AI development, especially if chip restrictions are circumvented.

The Race for AI Supremacy and Geopolitical Implications

The conversation turns to the intense competition between the U.S. and China in AI capabilities. Schmidt acknowledges that China has rapidly closed the gap, exemplified by the rise of DeepSeek, a Chinese AI model that recently surpassed other leading systems in intelligence benchmarks. This progress is partly due to innovative techniques like distillation and inference-time training, which allow smaller, more efficient models to perform at high levels.

This competitive landscape raises concerns about proliferation, control, and national security. Schmidt warns of a future where a handful of massive data centers—owned or controlled by governments—host the most powerful AI models, guarded with the utmost security akin to nuclear facilities. However, he also highlights the risk of AI capabilities proliferating to smaller, more accessible servers worldwide, especially through open-source models, which could empower bad actors, including terrorists.

The podcast explores the concept of “mutual AI malfunction,” a scenario analogous to nuclear deterrence, where the U.S. and China must carefully manage AI development to avoid catastrophic conflict. Schmidt advocates for transparency measures such as tracking chip locations and training runs to maintain stability. He stresses the urgency of initiating diplomatic and strategic dialogues now, drawing parallels to the pre-World War II era, to prevent unintended escalation.

Economic Transformation and the Future of Work

Turning to the economic impact, Schmidt offers a cautiously optimistic view of AI’s effect on jobs and productivity. He argues that, historically, automation has displaced the most dangerous and low-status jobs first, eventually creating new, higher-paying roles. AI, especially when paired with intelligent assistants, will augment human workers, enabling them to perform better and access more complex tasks.

He emphasizes that demographic trends, such as declining birth rates in many countries, make AI-driven productivity gains a national priority. In this context, AI adoption is not just an economic opportunity but a necessity to sustain growth and employment. Schmidt also highlights the importance of retraining and reskilling workers promptly, warning that delay could harm employees who face rapid technological change.

Education is a critical theme, with Schmidt lamenting the slow pace of innovation in teaching and advocating for AI-powered, gamified learning systems accessible in multiple languages. He believes young people, especially digital natives, will adapt quickly and thrive by applying AI to their interests, whether in climate science, energy, or other fields.

AI’s Impact on Creativity, Media, and Society

The discussion explores how AI is reshaping creative industries, including film and advertising. Schmidt envisions a future where individual creators, empowered by AI, can produce high-quality content that rivals traditional studios, though blockbuster productions will still involve human direction and collaboration. AI-generated voices and digital avatars will become increasingly realistic, raising questions about intellectual property and emotional impact.

Advertising and persuasion are also transformed by AI’s ability to tailor messages precisely to individuals, potentially making AI-driven persuasion more effective than human efforts. This raises concerns about misinformation, manipulation, and the health of democratic discourse. Schmidt reflects on the broader societal challenge of maintaining trust and shared values in an environment saturated with AI-generated content and personalized digital interactions.

Human Purpose and the Ethical Dimension

A profound part of the conversation addresses the risk of AI eroding human agency and purpose. Schmidt distinguishes between catastrophic scenarios popularized in fiction and a more subtle “drift” where humans might become overly reliant on AI, losing autonomy and the drive to overcome challenges. He stresses that purpose is fundamental to human well-being and that even as AI takes over many tasks, new challenges and complexities will arise, ensuring that human engagement remains vital.

The role of aesthetics and creativity gains prominence as AI becomes a force multiplier, enabling people to create virtually anything. Schmidt suggests that in an AI-augmented future, the question shifts from what we can create to why and how we choose to create, emphasizing meaning and intentionality.

Looking Ahead: The Path to Superintelligence

In closing, Schmidt reiterates that while specialized AI “savants” in various fields are essentially guaranteed within five years, the emergence of a unified superintelligence—an entity surpassing the collective human intellect—is still uncertain in timing and nature. The trajectory depends on breakthroughs in AI’s ability to transfer knowledge across domains, generate novel insights, and self-improve without human scaffolding.

He underscores the importance of vigilance, regulation, and international cooperation to harness AI’s benefits while mitigating risks.