Traditional, unilateral, centralized control is obsolete. When autonomous systems generate orders of magnitude more data than they can transmit, intelligence must live at the edge - and this constraint is revolutionizing everything from spacecraft to supply chains to healthcare.

William Van Dalsem, 42-year NASA veteran and Stanford adjunct lecturer, reveals why the future belongs to systems that think for themselves---not because it's elegant, but because physics demands it.

The Paradigm Shift:

→ The Edge Intelligence Imperative: Spacecraft orbiting Earth collect far more data than they can download---typically an order of magnitude difference. Factory sensors and autonomous vehicles face the same constraint. The bottleneck isn't computing power-it's bandwidth. Intelligence must live where decisions are made.

→ From "What" to "How": Organizations fail by conflating objectives with methods. Saying you need to "land on Mars using retro rockets" eliminates every methodological alternative you haven't imagined. Separate the destination from the journey.

→ The Modular Revolution: Van Dalsem's son built a state-of-the-art gaming computer from plug-and-play components---nearly supercomputer performance at home. What if spacecraft---or supply chains, or organizations---worked the same way? Standards enable innovation; vertical integration constrains it.

Ecosystem Impact:

→ Air traffic management evolved from one operator per aircraft to systems managing thousands of autonomous vehicles---the same pattern emerging in warehouse robotics, smart cities, and distributed manufacturing

→ Google's autonomous vehicles trained on moon-and-back distances (250,000 miles), capturing 90-99% of scenarios, yet still encounter situations they haven't seen - AI lacks mental models of physical reality. When confused, systems must "phone home," whether navigating streets

or diagnosing patients

→ The academia-industry-government "triad": diversity of perspective matters more than depth of expertise for solving novel problems

The Strategic Insight: Self-aware systems must be designed from inception, not retrofitted. Adding sensors to a Model T after it has been built isn't feasible. GE's digital transformation showed that "industrial equipment" must become "smart equipment" architecturally, not as an afterthought.

The Hidden Risk: LLMs hallucinate, lack context, and harm team dynamics when one "AI master" disconnects from collaborative processes. They're trained on historical data, embedding obsolete assumptions. Computational tools amplify, rather than replace, human judgment.

Strategic Reframe: Where must decisions be made, and what intelligence lives at the edge versus the center? Whether managing drone fleets, manufacturing networks, or distributed teams, resilient ecosystems distribute cognition across nodes rather than concentrating it in command centers.

The Van Dalsem Principle: When you specify both the "what" and the "how," you've eliminated every innovation you didn't imagine. Problem-focused innovation opens the aperture for solutions you might never imagine.

Guest: William Van Dalsem, Retired NASA Ames, Adjunct Lecturer, Stanford University

Host: Dyan Finkhousen, Founder & CEO, Shoshin Works

Ecosystemic Futures is a systems foresight series provided by Shoshin Works, evolved from our collaboration with NASA's Convergent Aeronautics Solutions Project.