Robust-Across-Futures Action
When the future is unknowable, choose actions that are ethical across a wide range of scenarios
- Difficulty
- Moderate
- Time to result
- ~ongoing to results
- Steps
- 3
- Confidence
- 80%
A decision framework for taking ethical action under deep uncertainty, developed by Ries for people paralyzed about what to do amid AI's wide 'aperture of possible futures.' Since no one can yet predict which future obtains, you pick actions today that make ethical sense across a wide variety of future scenarios, which breaks paralysis by surfacing a few clearly-good moves.
Origin
Eric Ries, extending Lean Startup's core stance on coping with uncertainty from entrepreneurship to ethical decision-making about AI and technology.
Core principles
- 01The aperture of possible AI futures is extremely wide and currently unknowable
- 02You cannot resolve the uncertainty yourself — you must wait for empirical research to narrow it
- 03Actions robust across many scenarios are the ones worth taking now
- 04Robustness analysis moves people from paralysis into concrete action
How to run it
- 1
Accept the uncertainty is real, not just personal
Recognize that not only do you not know the outcome, the most expert people don't either — they're still running the research to find out. Stop waiting to feel certain.
Watch out Firm convictions about which future is coming often drive drastic, contested actions that others believe cause the very dystopia you fear.
- 2
Enumerate plausible future scenarios
Lay out the wide range of futures the technology could produce, including opposing doomsday and utopia scenarios.
- 3
Select actions ethical across many of them
Choose the moves that make ethical sense in a wide variety of those scenarios and do those. These tend to be concrete and broadly agreed: more transparency, more capacity, more competent leaders, companies committed to clear values.
Pro tip Whatever your doomsday scenario is, more transparency right now is almost always among the robust actions.
Watch out Avoid one-scenario bets whose value collapses (or reverses) if a different future obtains, like drastic bans that others argue cause a worse outcome.
In the wild
Ries describes people having an existential crisis over what to do about AI, where every proposed action is called the 'evil' one by someone. Applying the robust-across-futures lens surfaces clear shared actions — transparency, capacity, competent leadership, values-committed companies — that hold regardless of which future arrives.
→ People move out of total paralysis into taking worthwhile action without needing to predict the future.
Common mistakes
Acting on a firm conviction about one future
Everyone's certainty differs, and for every drastic move (e.g. banning open source) someone credibly argues it causes the dystopia instead of preventing it.
Is it for you?
Best for
Founders and operators deciding how to act ethically amid AI or other high-uncertainty technology shifts
Not ideal for
Low-uncertainty decisions where the outcome space is narrow and well understood and a targeted bet is simply better
From the transcript
“pick actions today that make ethical sense in a wide variety of future scenarios and do those”
“whatever your doomsday scenario is we need more transparency right now”
“how do you take ethical action in the face of high uncertainty”
From the episode
Reflections on a movement
Eric Ries (creator of the Lean Startup methodology)