Whole-System Redesign Method
Replace isolated fixes with coordinated changes that create a new equilibrium
- Difficulty
- Expert
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 94%
Hillis distinguishes point optimization from redesigning the arrangement that joins the points. Start by mapping everything required for the outcome, including elements that are easy to overlook because they sit outside the visible product. In agriculture, that includes weather prediction, fertilizer, labor, greenhouse architecture, plant varieties, refrigerated transport, and soil microbiomes—not merely planting and harvesting. Mature fields may have strong solutions at each point while the total system remains inefficient. The opportunity appears when several constraints can be changed together to create a different stable configuration, or equilibrium, that no single improvement could reach. Because coordinated redesign is complex and commercially awkward, the method also requires a visionary operator or funder willing to support the transition.
Origin
Hillis explained the method through Applied Invention's work on regional agriculture and contrasted system redesign with commercially attractive point solutions.
Core principles
- 01Mature systems often contain well-optimized parts but poor overall arrangements
- 02The real system extends beyond the most visible production step
- 03Some improvements work only when several constraints change together
- 04A new equilibrium can unlock value unavailable to isolated fixes
- 05System projects need patient, visionary support
How to run it
- 1
Define the end-to-end outcome
State the result the system must deliver and trace the path from raw inputs to the end user.
Pro tip Follow the product beyond the boundary of the obvious industry process.
Watch out A narrow boundary will hide the leverage.
- 2
Map hidden dependencies
Add labor, energy, capital, logistics, information, regulation, and environmental constraints.
Pro tip Ask what must happen even though customers never see it.
Watch out Do not assume the named industry contains the whole system.
- 3
Audit point optimization
Identify components that specialists have already improved and locate inefficiencies created by how those components interact.
Pro tip Look for locally rational decisions that produce a poor global result.
Watch out Reinventing a well-solved component adds complexity without leverage.
- 4
Bundle coordinated changes
Specify which constraints must change simultaneously for an alternative arrangement to become viable.
Pro tip Describe causal dependencies between changes rather than listing ideas.
Watch out Changing only one interdependent variable may make the system worse.
- 5
Model the new equilibrium
Explain how the revised components reinforce one another and where the system settles once the transition is complete.
Pro tip Test labor, cost, and operating assumptions together.
Watch out A desirable endpoint is not enough if no feasible transition reaches it.
- 6
Secure visionary backing
Find an operator, patron, or funder who understands the system-level opportunity and accepts coordinated risk.
Pro tip Pair technical ambition with a real economic opportunity created by solving the problem.
Watch out Conventional project finance may reject the work before the combined case is visible.
In the wild
Growing better food near Boston could not be solved by moving a farm alone. Hillis described changing labor use, cold-climate greenhouses, plant varieties, greenhouse architecture, worker jobs, and soil microbiomes together so crops no longer needed to survive shipping for thousands of miles.
→ The coordinated changes define another viable equilibrium in which many crops can be grown much closer to consumers.
Common mistakes
Optimizing the visible point
Improving one task can miss constraints elsewhere in the supply chain that determine the total outcome.
Changing one coupled variable
A new equilibrium may require simultaneous changes to labor, infrastructure, biology, and logistics.
Ignoring transition support
System redesign is more likely to fail without a funder or operator willing to carry cross-component risk.
Is it for you?
Best for
Complex industries whose labor, infrastructure, logistics, and products constrain one another.
Not ideal for
Problems with a proven isolated fix or teams unable to coordinate multiple dependencies.
From the transcript
“are there ways to change the system rather than solving individual Point problems within the system”
“people have Point optimized out most of the specific Solutions done a pretty good job of that but very few if anybody I've kind of…”
“if you change all of those things up at once there's another equilibrium point a Nash equilibrium where there's another sweet spot of things working…”
From the episode
#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
Danny Hillis