MIT Professor on AI: "We Have 18 Months Before It's Game Over"

What happens when artificial intelligence stops being something we use and starts becoming something that shapes who we are?

Julian sits down with Ramesh Raskar for a hopeful, wide-ranging conversation about where AI is actually headed and where it should go. Instead of leaning into the usual doomsday fears, they talk about a different risk, a future where intelligence is centralized, creativity is squeezed out, and people slowly lose their sense of purpose.

They discuss what an “agentic” future could look like, one where everyone has their own AI agent working on their behalf, not owned by a company but shaped by personal values, goals, and lived experience. They talk about how this could transform healthcare, education, and work, not by replacing humans, but by amplifying what makes human contribution meaningful. There’s a strong emphasis on dignity, pride in craft, and why losing those things could be more damaging than job loss itself.

Ramesh Raskar is an MIT professor, researcher, and inventor whose work spans artificial intelligence, imaging, robotics, and decentralized systems. He has led research teams at Google, Facebook, and Apple, and focuses on building human-centered AI technologies designed to support shared prosperity.

What You’ll Learn: Why centralized AI risks reducing human agency and purpose How decentralized AI agents could reshape work, education, and healthcare The difference between efficiency-driven AI and human-centered intelligence Why loss of meaning may be a bigger threat than job displacement How shared prosperity depends on ownership, not just access What an “agentic society” could look like in practice Why AI literacy is essential for the decade ahead How adaptability and learning-to-learn shape human flourishing

⏱ Timestamps 00:00 Intro 00:51 How Can AI Benefit Society? 02:40 The Real Risk of AI Isn’t Doomsday 05:45 From Using AI to Owning AI 10:16 Why Centralized AI Fails 12:25 The Rise of Personal AI 15:10 When Work Loses Meaning 17:38 Why Purpose Still Matters 20:13 A World Where Everyone Has an AI Agent 25:42 Inside an Agentic Society 33:11 How Power and Bias Scale 36:41 Why Decentralization Works 39:07 Lessons From the Internet 44:07 What the World Could Look Like in 2030 47:20 How AI Could Change Healthcare 48:45 Collective Intelligence at Scale 53:55 Who Wins When Data Is Decentralized 56:45 How AI Doomsday Actually Happens 01:01:44 Why Phase One Was Necessary 01:04:20 Why AI Literacy Matters 01:07:20 Learning How to Learn 01:08:22 What Kids Should Learn Today 01:10:05 A Simple Rule for the Future

Want to Follow Ramesh's Mission? Join here: https://nanda.mit.edu, projectnanda.org, join39.org LinkedIn: https://www.linkedin.com/in/raskar MIT Media Lab: https://www.media.mit.edu/people/raskar/overview/

Connect with Julian 📸 Instagram: @thebeyondtomorrowpodcast 🌐 Website: www.beyondtomorrowpodcast.com 📩 Email: [email protected] 🎧 Subscribe on Spotify, Apple, and YouTube for more stories from the edge of what is possible.