TThe Tim Ferriss Show
← All frameworks
StrategyDanny Hillis

Four-Gate Non-Redundant Project Filter

Choose projects that excite, sustain, teach, and would otherwise go undone

Difficulty
Moderate
Time to result
~weeks to results
Steps
5
Confidence
99%

Applied Invention filters candidate projects through distinct gates rather than letting enthusiasm make every decision. First, a senior partner must be genuinely excited, usually because the work could have major impact or uses unusually compelling technology. Second, partners who are less emotionally attached evaluate whether the financial model is sustainable, including bounded downside or portfolio-level upside. Third comes Hillis's non-redundancy criterion: reject work that somebody else will do anyway, and identify the unique reason this team is necessary. Hillis later articulated a fourth recurring consideration—whether the project creates an opportunity to learn something important. Keeping enthusiasm, financial scrutiny, uniqueness, and learning separate reduces motivated reasoning while preserving ambitious projects that conventional companies would overlook.

Origin

Hillis described the formal selection discipline used by the partners at Applied Invention and added the learning gate during the interview.

Core principles

  • 01A senior owner must care deeply about the project
  • 02Financial judgment should come from someone less emotionally invested
  • 03Work is wasteful when others will do it adequately anyway
  • 04A project should create a meaningful learning opportunity
  • 05Sustainability matters even when wealth maximization does not

How to run it

  1. 1

    Assign an excited owner

    Require a senior partner to explain why the project matters and commit genuine attention to it.

    Pro tip Separate excitement about impact from fascination with technology.

    Watch out Do not proceed when enthusiasm is merely polite or delegated.

  2. 2

    Run an independent sustainability test

    Ask partners who are not the project's champions to evaluate how it pays for itself, limits losses, or contributes to a balanced portfolio.

    Pro tip Let skeptics own this gate so excitement cannot grade itself.

    Watch out Sustainability does not require maximizing profit, but it does require a credible way to continue.

  3. 3

    Apply non-redundancy

    Ask whether the desired outcome will happen anyway and why your team changes its timing, quality, or existence.

    Pro tip Look for ignored assumptions, disciplinary gaps, or barriers that explain why others have not acted.

    Watch out A profitable opportunity is still redundant if others will deliver it adequately without you.

  4. 4

    Specify the learning return

    Identify the field, mechanism, or capability the work will let the team understand.

    Pro tip Prefer learning that strengthens the project's impact rather than serving as a distraction.

    Watch out Curiosity alone cannot rescue a project that fails the other gates.

  5. 5

    Approve only the complete case

    Proceed only when excitement, sustainability, non-redundancy, and learning are all explicit enough to defend.

    Pro tip Write each gate's evidence before committing resources.

    Watch out Do not blend the gates into one vague overall score.

In the wild

Parallel computing

Hillis saw that multicore silicon, compiler technology, digital cameras, and digitized audio had created the preconditions for parallel computing. Prevailing belief still treated the approach as impossible, while the human brain gave him reason to believe parallel processing could work for AI.

Thinking Machines built the first large parallel computer and produced the world's fastest computers for years, though the business itself was not managed sustainably.

Zero-trust packet routing

Hillis gathered early internet experts to ask how Internet Protocol would have been designed with modern cybersecurity knowledge. The foundational redesign was unlikely to be attempted by a normal commercial company, but the team found it could be deployed as an overlay on the existing internet.

They developed zero-trust packet routing, and Oracle later announced a cloud product using the protocol.

Common mistakes

Letting champions grade economics

The most excited person is poorly positioned to provide independent financial discipline.

Confusing profit with necessity

A project can make money and still waste the team's time if it will happen anyway.

Ignoring sustainability

Hillis linked Thinking Machines' failure to not treating financial sustainability as seriously as the technical work.

Is it for you?

Best for

Inventors, venture studios, and research teams choosing among many plausible opportunities.

Not ideal for

Mandatory operational work where the project cannot be declined.

From the transcript

we put three tests on things

Danny Hillis · 31:30

if somebody else is going to do it why should you do it you're wasting your time

Danny Hillis · 33:00

there's always something that you kind of want to learn about

Danny Hillis · 1:01:00

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