Download Column: The Future of Nuclear Power and the Social Media AI Boom

Introduction: A Must-Read from Tech Daily

The Download is a midweek newsletter from MIT Technology Review, curating the latest tech developments daily to bring readers the pulse of the tech world. This edition, written by Rhiannon Williams and published on February 4, 2026, focuses on two themes: the future of nuclear power plants and the social media-driven AI frenzy. As the AI revolution sweeps the globe, these issues cut to the core of energy bottlenecks and public opinion bubbles.

AI is driving unprecedented investment in data centers and the energy supply needed to support their massive computing demands. One potential source of that power...

Why Are AI Companies Betting on Next-Generation Nuclear Power?

The rise of AI is reshaping the global energy landscape. Training large language models like GPT-4 requires immense computing resources, which rely on massive data centers. According to the International Energy Agency (IEA), global data center electricity consumption is expected to account for 8% of total electricity use by 2030, equivalent to the level of Japan. Traditional power grids struggle to meet this explosive demand. Renewable sources like wind and solar, while clean, are intermittent due to weather and cannot provide 24/7 stable power.

Nuclear energy is making a strong comeback in a new form. Tech giants such as Microsoft, Google, and Amazon are heavily investing in small modular reactors (SMRs) and advanced nuclear reactors. These technologies are smaller, faster to build, and safer, and can be deployed modularly next to data centers, avoiding long-distance transmission losses. For example, Microsoft has signed an agreement with Helion Energy to bring its first fusion reactor online by 2028; Google has partnered with Kairos Power, investing hundreds of millions of dollars to develop molten salt-cooled reactors. OpenAI founder Sam Altman has also invested in Oklo to commercialize micro-nuclear reactors.

The nuclear renaissance is not a sudden whim. Historically, accidents at Three Mile Island (1979) and Chernobyl (1986) cast a shadow over nuclear power, leading to a sharp decline in new projects globally. But fourth-generation nuclear technology addresses these pain points: passive safety systems can cool without human intervention, fuel utilization increases by 90%, and waste is significantly reduced. China is a leader in this field, having built the world's first fourth-generation nuclear power plant, Hualong One, and exporting it to multiple countries. In 2023, global nuclear investment rebounded to $50 billion, with AI demand contributing 30%.

Editor's Note: The marriage of AI and nuclear power is a win-win, but challenges remain. Regulatory approvals take years, high upfront costs (around $6,000 per kilowatt) test corporate patience, and public concerns about nuclear safety need to be addressed through education. China's experience offers lessons: through policy support and localized supply chains, its installed nuclear capacity ranks first in the world. If AI giants succeed, this could usher in a "Nuclear 2.0" era, advancing carbon neutrality goals.

How Social Media Ignites the AI Frenzy

In contrast to energy realities, the narrative on social media about AI is feverish. Twitter (now X), TikTok, and Reddit are flooded with AI-generated art, chatbot demos, and predictions that "AI will replace humans." The viral spread of Midjourney and Stable Diffusion allows ordinary users to create masterpieces with a click, garnering millions of likes and shares. In 2025, the "AI bro" meme swept the globe, investors rushed to buy NVIDIA stock, and its market capitalization soared to $4 trillion.

This frenzy has its roots: social algorithms favor surprising content, and AI demo videos are easily viral. The launches of Elon Musk's xAI and Anthropic's Claude model are often accompanied by tweet storms, amplifying exposure. But signs of a bubble are emerging. AI startups raised over $100 billion in funding in 2024, yet many are still proof-of-concept, with actual deployment rates below 20%. The proliferation of deepfakes exacerbates misinformation, and the EU has introduced the AI Act to regulate high-risk applications.

Social media not only amplifies but also shapes public perception. Surveys show that 65% of Gen Z learn about AI from TikTok rather than academic reports. This drives talent inflows: Harvard dropouts found AI companies, and hiring wars in Silicon Valley double salaries. But experts warn that overhype could lead to an "AI Winter 2.0," similar to the expert system bubble in the 1980s.

Editor's Note: The social media AI frenzy is a double-edged sword. On one hand, it sparks innovation and democratizes technology; on the other, it creates FOMO (fear of missing out) and misleads investments. The media should provide balanced coverage, emphasizing the gradual nature of AI rather than sci-fi visions. Chinese platforms like Bilibili, with their AI tutorials, are educating the public in an accessible way, avoiding blind follow-the-trend behavior.

Looking Ahead: Balancing Energy and Public Opinion in the AI Era

This issue of The Download reveals the dual face of AI: energy hunger and public opinion frenzy. The revival of nuclear power may address the urgent need, but requires global cooperation; the AI frenzy needs rational cooling to be sustainable. Tech giants are accelerating their actions, and policymakers must follow suit. Readers are encouraged to watch for further developments—perhaps the next breakthrough is just around the corner.

This article is compiled from MIT Technology Review.