Loading...
December 11, 2024 • 2hr 36min
#782: Legendary Inventor Danny Hillis (Plus Kevin Kelly) — Unorthodox Lessons from 400+ Patents, Solving the Impossible, Real Al vs. “AI”, Hiring Richard Feynman, Working with Steve Jobs, Creating Parallel Computing, and Much More
The Tim Ferriss Show
Danny Hillis is an inventor, scientist, author and engineer who pioneered parallel computers while at MIT. He founded Thinking Machines Corporation, served as Disney's first Fellow, and is now a founding partner at Applied Invention. The conversation explores his unique approach to learning and invention, his views on technology and consciousness, and his work on projects ranging from cancer diagnostics to the 10,000 Year Clock.
Early Career and Education (14:23)
While at MIT, Hillis recognized that AI would require powerful parallel computers that didn't yet exist. He started building one as a graduate student, eventually founding a company despite MIT's initial objections. The company successfully built the world's fastest computers but struggled as a business.
Key insight: Technical success doesn't automatically translate to business success
Learning approach: Found creative ways to access experts and resources, like studying DARPA proposals in MIT's lobby
Built relationships through demonstrating capability and asking thoughtful questions
The Disney Years (19:17)
After Thinking Machines, Hillis took a position as Disney Fellow and VP of Imagineering - his only traditional job. This experience provided crucial lessons in storytelling and artistic approaches to problem-solving.
Engineering vs. Artistic mindset:
Engineers focus on technical functionality
Artists/storytellers focus on emotional experience and narrative
Key lesson: "The story and how people relate to something is often more important than the technical details"
Approach to Learning and Innovation (47:38)
Hillis describes his method for entering new fields and pursuing innovations:
Do homework first to identify interesting questions
Find and learn from experts who are willing to engage with "dumb" questions
Look for preconditions where pieces are in place but haven't been connected
Focus on systemic changes rather than point solutions
The Three Criteria for Projects (1:06:34)
At Applied Invention, projects must meet three key criteria:
Excitement: A senior partner must be genuinely passionate about it
Financial viability: Must have a sustainable model
Non-redundancy: Won't happen otherwise
"The non-redundancy criterion is the hardest thing to do...you would say no if it's going to happen anyway"
Views on AI and Intelligence (1:31:28)
Hillis offers unique perspectives on AI and intelligence:
Current AI is primarily good at imitation - like a child's first learning stage
Intelligence is multi-faceted - what we thought was hard (chess) turned out to be easy; what seemed easy (vision) proved difficult
Human intelligence may be just one small corner of possible types of intelligence
"We're short-term overestimating capabilities but long-term underestimating what we'll accomplish"
The Entanglement of Natural and Artificial (1:46:23)
Hillis describes how the boundaries between natural and artificial systems are blurring:
Technology becoming more nature-like:
No one can fully understand or control modern systems like the internet
We rely more on "magic incantations" than deep understanding
Natural systems becoming more engineered:
Atmosphere, genes, minds becoming technological artifacts
Distinction between natural and artificial may eventually disappear
The 10,000 Year Clock (2:27:22)
One of Hillis's most ambitious projects is a mechanical clock designed to run for 10,000 years:
Purpose: Create a symbol that helps people think about the long-term future
Impact through story: The myth and story of the clock may be as important as the physical artifact
Design philosophy: Focused on visitor experience and emotional impact rather than just technical function
"Stories are actually what really lasts...an idea has a lot more sticking power than any physical thing you could build"
#782: Legendary Inventor Danny Hillis (Plus Kevin Kelly) — Unorthodox Lessons from 400+ Patents, Solving the Impossible, Real Al vs. “AI”, Hiring Richard Feynman, Working with Steve Jobs, Creating Parallel Computing, and Much More
0:00
-0:00
Key Takeaways
- Learning from experts: Danny Hillis emphasizes the importance of doing homework before meeting experts, asking thoughtful questions that make them think, and learning primarily through conversations rather than just reading papers
- Three criteria for pursuing projects:
- A senior partner must be genuinely excited about it
- It needs to make financial sense or be sustainable
- It must be non-redundant - something that wouldn't happen otherwise
- Systems thinking is critical - looking at how all components interact rather than just optimizing individual point solutions
- The distinction between natural and artificial systems is becoming increasingly blurred - technology is becoming more nature-like and natural systems are becoming more engineered
- On consciousness and AI: Consciousness may be less central to intelligence than we think - it could just be a "hack" that allows us to access our own thoughts
Introduction
Danny Hillis is an inventor, scientist, author and engineer who pioneered parallel computers while at MIT. He founded Thinking Machines Corporation, served as Disney's first Fellow, and is now a founding partner at Applied Invention. The conversation explores his unique approach to learning and invention, his views on technology and consciousness, and his work on projects ranging from cancer diagnostics to the 10,000 Year Clock.
Early Career and Education (14:23)
While at MIT, Hillis recognized that AI would require powerful parallel computers that didn't yet exist. He started building one as a graduate student, eventually founding a company despite MIT's initial objections. The company successfully built the world's fastest computers but struggled as a business.
- Key insight: Technical success doesn't automatically translate to business success
- Learning approach: Found creative ways to access experts and resources, like studying DARPA proposals in MIT's lobby
- Built relationships through demonstrating capability and asking thoughtful questions
The Disney Years (19:17)
After Thinking Machines, Hillis took a position as Disney Fellow and VP of Imagineering - his only traditional job. This experience provided crucial lessons in storytelling and artistic approaches to problem-solving.
- Engineering vs. Artistic mindset:
- Engineers focus on technical functionality
- Artists/storytellers focus on emotional experience and narrative
- Key lesson: "The story and how people relate to something is often more important than the technical details"
Approach to Learning and Innovation (47:38)
Hillis describes his method for entering new fields and pursuing innovations:
- Do homework first to identify interesting questions
- Find and learn from experts who are willing to engage with "dumb" questions
- Look for preconditions where pieces are in place but haven't been connected
- Focus on systemic changes rather than point solutions
The Three Criteria for Projects (1:06:34)
At Applied Invention, projects must meet three key criteria:
- Excitement: A senior partner must be genuinely passionate about it
- Financial viability: Must have a sustainable model
- Non-redundancy: Won't happen otherwise
- "The non-redundancy criterion is the hardest thing to do...you would say no if it's going to happen anyway"
Views on AI and Intelligence (1:31:28)
Hillis offers unique perspectives on AI and intelligence:
- Current AI is primarily good at imitation - like a child's first learning stage
- Intelligence is multi-faceted - what we thought was hard (chess) turned out to be easy; what seemed easy (vision) proved difficult
- Human intelligence may be just one small corner of possible types of intelligence
- "We're short-term overestimating capabilities but long-term underestimating what we'll accomplish"
The Entanglement of Natural and Artificial (1:46:23)
Hillis describes how the boundaries between natural and artificial systems are blurring:
- Technology becoming more nature-like:
- No one can fully understand or control modern systems like the internet
- We rely more on "magic incantations" than deep understanding
- Natural systems becoming more engineered:
- Atmosphere, genes, minds becoming technological artifacts
- Distinction between natural and artificial may eventually disappear
The 10,000 Year Clock (2:27:22)
One of Hillis's most ambitious projects is a mechanical clock designed to run for 10,000 years:
- Purpose: Create a symbol that helps people think about the long-term future
- Impact through story: The myth and story of the clock may be as important as the physical artifact
- Design philosophy: Focused on visitor experience and emotional impact rather than just technical function
- "Stories are actually what really lasts...an idea has a lot more sticking power than any physical thing you could build"
Conclusion
Danny Hillis exemplifies a unique approach to innovation that combines technical expertise with storytelling, systems thinking, and a focus on long-term impact. His work spans multiple fields and demonstrates the value of crossing disciplinary boundaries while maintaining a clear framework for project selection. His perspectives on AI, consciousness, and the future of technology offer valuable insights for understanding our rapidly evolving world.