What Is the Real World AI Stage?

While TechCrunch Disrupt has long covered AI software and enterprise workflows on its AI Stage (presented by Google Cloud), the new Real World AI Stage examines what happens when AI leaves the data center and enters the physical world. The stage covers autonomous hardware in defense, manufacturing, and industrial operations, as well as edge AI deployments where cloud connectivity is unavailable, and the deep challenges of moving from prototype to scalable production . One session even tackles the use of AI in computational biology to revive extinct species .

Confirmed Sessions and Speakers

TechCrunch has announced four sessions for the Real World AI Stage, with speaker additions expected as the event approaches .

1. Building AI Systems When Failure Is Not an Option

• Speaker: Nate Michael, CTO, Shield AI
• Focus: This session addresses safety-critical AI deployment in autonomous vehicles, defense platforms, and industrial systems where physical failure has severe consequences. Key topics include creating a safety culture in hardware organizations, AI validation and testing methods, navigating regulation, and building trust in high-risk environments .

2. Can We Engineer Nature’s Comeback?

• Speaker: Ben Lamm, CEO, Colossal Biosciences
• Focus: A fireside chat exploring technologies used to revive extinct species, the role AI plays in modern biology, and the ethics of engineering nature. Colossal Biosciences has turned de-extinction from science fiction into a billion-dollar business, and the session will debate whether it represents a breakthrough for conservation or a distraction from protecting existing species .

3. Working at the Edge: How AI Operates When the Cloud Is Unavailable

• Panelists: Dr. Ali Agha (CEO/Founder, FieldAI), Michelle Lee (CEO/Founder, Medra), and Aidan Madigan-Curtis (Partner, Eclipse Ventures)
• Focus: Edge AI deployment in environments without reliable cloud connectivity. The panel examines architectures that push inference onto low-power ASICs and ruggedized GPUs, enabling local millisecond-level decisions for robotics, field operations, and remote infrastructure .

4. From Prototype to Production: Why Most Deep Tech Startups Fail

• Panelists: John Mackey (CEO, MBRYONICS), Boris Sofman (CEO/Co-founder, Bedrock Robotics), and Adrian Macneil (CEO, Foxglove)
• Focus: Scaling challenges for deep-tech hardware and robotics companies transitioning from prototype to manufacturing — often called the "valley of death" for physical tech startups. Covers supply chain, unit economics, manufacturing partnerships, and organizational scaling .

All Six Stages at Disrupt 2026

The Real World AI Stage is one of six parallel tracks. According to the official TechCrunch Disrupt 2026 agenda, the full conference features :

Why a Separate Stage for Physical AI?

TechCrunch’s decision to split AI into two dedicated stages signals that AI robotics, edge inference, and autonomous hardware have moved from niche topics to the mainstream of the startup and investment landscape . The Real World AI Stage reflects a growing recognition that building safe, reliable, and scalable AI for the physical world involves fundamentally different challenges than training large language models — validation, safety certification, manufacturing supply chains, and edge deployment constraints are now front and center for investors and founders alike .

What Is the Real World AI Stage?

While TechCrunch Disrupt has long covered AI software and enterprise workflows on its AI Stage (presented by Google Cloud), the new Real World AI Stage examines what happens when AI leaves the data center and enters the physical world. The stage covers autonomous hardware in defense, manufacturing, and industrial operations, as well as edge AI deployments where cloud connectivity is unavailable, and the deep challenges of moving from prototype to scalable production . One session even tackles the use of AI in computational biology to revive extinct species .

Confirmed Sessions and Speakers

TechCrunch has announced four sessions for the Real World AI Stage, with speaker additions expected as the event approaches .

1. Building AI Systems When Failure Is Not an Option

• Speaker: Nate Michael, CTO, Shield AI
• Focus: This session addresses safety-critical AI deployment in autonomous vehicles, defense platforms, and industrial systems where physical failure has severe consequences. Key topics include creating a safety culture in hardware organizations, AI validation and testing methods, navigating regulation, and building trust in high-risk environments .

2. Can We Engineer Nature’s Comeback?

• Speaker: Ben Lamm, CEO, Colossal Biosciences
• Focus: A fireside chat exploring technologies used to revive extinct species, the role AI plays in modern biology, and the ethics of engineering nature. Colossal Biosciences has turned de-extinction from science fiction into a billion-dollar business, and the session will debate whether it represents a breakthrough for conservation or a distraction from protecting existing species .

3. Working at the Edge: How AI Operates When the Cloud Is Unavailable

• Panelists: Dr. Ali Agha (CEO/Founder, FieldAI), Michelle Lee (CEO/Founder, Medra), and Aidan Madigan-Curtis (Partner, Eclipse Ventures)
• Focus: Edge AI deployment in environments without reliable cloud connectivity. The panel examines architectures that push inference onto low-power ASICs and ruggedized GPUs, enabling local millisecond-level decisions for robotics, field operations, and remote infrastructure .

4. From Prototype to Production: Why Most Deep Tech Startups Fail

• Panelists: John Mackey (CEO, MBRYONICS), Boris Sofman (CEO/Co-founder, Bedrock Robotics), and Adrian Macneil (CEO, Foxglove)
• Focus: Scaling challenges for deep-tech hardware and robotics companies transitioning from prototype to manufacturing — often called the "valley of death" for physical tech startups. Covers supply chain, unit economics, manufacturing partnerships, and organizational scaling .

All Six Stages at Disrupt 2026

The Real World AI Stage is one of six parallel tracks. According to the official TechCrunch Disrupt 2026 agenda, the full conference features :

Why a Separate Stage for Physical AI?

TechCrunch’s decision to split AI into two dedicated stages signals that AI robotics, edge inference, and autonomous hardware have moved from niche topics to the mainstream of the startup and investment landscape . The Real World AI Stage reflects a growing recognition that building safe, reliable, and scalable AI for the physical world involves fundamentally different challenges than training large language models — validation, safety certification, manufacturing supply chains, and edge deployment constraints are now front and center for investors and founders alike .