Thanks for this Andrew. Very confirmation of what we doing at PreEmpt.Life, using strategic foresight to PreEmpt Disaster.
WHAT WE ARE BUILDING
PreEmpt Disaster — a three-layer temporal intelligence system:
Historical layer — global database of mass harm events, Harm Gap calculated retrospectively across the full record
Standing layer — current hazard estimates for known high-risk zones, updated continuously
Live layer — ALEXIS daily alerts detecting pre-event signals in real time
Incorporates and extends the existing EMS. Public good. Built on EM-DAT, GDACS, NOAA, Munich Re, Swiss Re, Sendai Framework Monitor, plus original data collection including lessons learned and after-action reports.
THE METHODOLOGICAL FOUNDATION
WHY-54 (Hazard Decision Intelligence) ratified 25 June 2026. Three components: Standing Hazard Estimate / Resilience Coefficient / Post-Event Actuals Comparison. The Harm Gap = pre-event modelled harm minus post-event actual harm. Positive = Resilience Dividend. Negative = Preparedness Debt. Counterfactual Cost Ratio and Decision Attribution Layer registered. Cross-hazard: geophysical, climate, pandemic, conflict, industrial. Master Index v4.3.
THE WHITE PAPER
Title: The Price of Knowing
Subtitle: Why the cost of ignored hazard always exceeds the cost of preparedness — and how to measure the gap before the event, not after.
Status: Section 1 drafted. Architecture complete — ten sections. Venezuela 2026 as opening case. Primary audience: emergency response organisations and organisations in affected areas.
The white paper is now the introduction to PreEmpt Disaster — not a standalone publication.
THE INTELLECTUAL SPINE — THREE AUTHORS, ONE ARGUMENT
Gill Kernick — Catastrophe and Systemic Change (2026)
The Grenfell Tower fire and dozens of other disasters examined. Her central finding: the system is perfectly designed to ensure we do not learn. Her key distinction: piecemeal change fixes a part; systemic change shifts the conditions holding the system in place. Her six diagnostic questions for systems that don't want to change are the analytical instrument for the Decision Attribution Layer.
Most importantly for the white paper: she establishes that the Grenfell deaths were foretold — in coroners' reports, in MPs' letters, in residents' emails, in cladding contractors' own communications. The word foretold is now active PreEmpt Disaster vocabulary. Use it deliberately and consistently.
Bazerman & Watkins — Predictable Surprises (Harvard Business School, 2004)
Core definition: a predictable surprise arises when leaders had all the data and insight needed to recognize the potential for, even the inevitability of, a crisis — and failed to act. Three root causes of why the foretelling is ignored: cognitive bias, organizational failure, political interest. Five reasons we are most likely to be surprised — positive illusions, status quo bias, minority interests benefiting from inaction, lack of vividness, institutional processes that suppress emerging threats.
Prevention framework: Recognition / Prioritization / Mobilization. Critical finding for PreEmpt Disaster: organizations achieve the greatest success preventing predictable surprises by adopting blanket measures across a spectrum of disasters — not addressing them one at a time. This validates cross-hazard universality directly.
The combined argument in one sentence:
Disasters are foretold. The foretelling is systematically suppressed — by cognitive bias, institutional failure, and political interest. PreEmpt Disaster is the infrastructure that makes the foretelling impossible to ignore — because it attaches a number to it before the event, not after.
That is the sentence the new session opens with.
THE LANGUAGE REGISTER
New vocabulary now active — use consistently throughout:
Foretold — the disaster was announced in advance, in the record, before it happened
Predictable surprise — Bazerman & Watkins' term, now part of PreEmpt Disaster's intellectual warrant
Harm Gap — the measurable distance between what was modelled and what occurred
Resilience Dividend — positive gap: the society built something
Preparedness Debt — negative gap: identifiable actors made identifiable choices
Counterfactual Cost Ratio — what preparedness cost versus what inaction cost
Decision Attribution Layer — the trail from harm back to decision
Systemic change — Kernick's frame: shifting the conditions holding the system in place, not patching individual failures
The problem I have with this outline is that it perpetuates many of the same controllability fallacies that lead to these collapses/disasters in the first place.
Engineering systems with more data isn't sufficient for a complex system that's teetering on a phase-shift collapse into a domain where there is no useful data.
The systems shift has to begin with not knowing and experimenting into the unknown over itemizing all knowns and engineering to them.
You're pointing at something real, and we haven't resolved it visibly enough in the framing.
The Harm Gap is not a prediction instrument. It measures the distance between what was knowable — seismic hazard, population exposure, governance quality — and what was done about it. Venezuela's fault system was not unknown. The governance capacity to respond was measurable. What failed was not knowledge. It was the decisions made, or not made, in the window between knowing and the event. That's the complicated-to-complex failure mode PreEmpt Disaster is designed to make visible and attributable.
You're right that a different failure mode exists — the genuine phase shift where prior data misleads rather than informs. We don't claim the system handles that yet. But we do have a working test model of experimenting with the unknown The live crisis layer has an explicit chaotic domain protocol whose instruction is: stabilise before you analyse. Not more data. Hold the line on values while the picture forms. That logic should be visible at the PreEmpt Disaster level, not buried in operational detail — and you've identified a real gap.
Where we'd push back: the binary between data engineering and experimenting into the unknown is false in the specific geography this system operates in. People are not dying because their governments lacked a probe-sense-respond epistemology. They are dying because knowledge that existed — often for decades — was ignored, suppressed, or never reached those with authority to act. The Harm Gap measures that distance. The Decision Attribution Layer names the decisions that produced it. That is a governance accountability instrument for the complicated domain. Not a controllability claim.
If you're working on what happens when decision support hits genuine phase-shift territory, that's a conversation we want to have.
I find the question, "How to work with systems that don’t want to change?" interesting, because systems are systems precisely because they have continuity. Without continuity, we have chaos.
And we shouldn't undervalue the value of continuity, of predictability, of something that we know or we're at least familiar with, versus the unknown, untested, and not-yet-experienced.
As Diane Musho Hamilton and folks working with Polyvagal Theory have noted, sameness calms our nervous systems while difference excites our nervous systems. People are encountering so much difference now, so quickly, that it's overwhelming our nervous systems. We don't have adequate practices to deal with that. And since so many of us have dysregulated, even traumatized, nervous systems, and predictability and repetition are natural salves for this, we tend to prefer the devil we know to the devil we don't, even if we literally believe it's a devil.
Whether we are for a system or against a system, whether we benefit from it or we are harmed by it, we all get something from it. At the very least, even if I resist the system, it gives my life some meaning and direction. So we all carry water for dominant systems, even if we're completely dissatisfied with them.
William James wrote in 1890 (in The Principles of Psychology), "The great thing in all education is to make our nervous system our ally instead of our enemy". That's the one element I'm not seeing in the Grenfell Model of Change, and maybe one we most need.
Hi Russ, One of my systems heuristics is Donald Schon’s principle of ‘dynamic conservatism’, which says that in response to an external shock a system will adjust by the smallest amount required to absorb the shock. When I’m explaining three horizons to people, I describe the 1st Horizon as the land of the “maintainer”, rather than the “manager” (as in much of the literature), because we need people who are interested in the integrity and stability of the current system. But it depends what the purpose of the present system is. When you read Gill Kernick’s book, or Peter Abbs’ coverage of the Grenfell Fire, it’s pretty clear that the purpose of the systems around Grenfell were either to maximise profit without regard for safety (in the case of the cladding contractors) or to minimise cost without regard for the interests of the residents (in the case of the TMO). Clearly these two sub-systems align quite strongly, and any notion of public interest is lost: the contractors were helped by regulatory collusion. In the terms of Charles Hampden-Turner’s Dilemma Resolution model, the Rock—the current system structure—gets everything it wants in the short term, but not in the long-term, precisely because it is not able to absorb the change elements that are represented by the Whirpool. Would it help you if the question were reframed as “How to work with systems that need to be able change but are unable to because they are locked into producing malign outcomes?”
I sometimes invite people to consider questions like, "What's keeping this system in place?" or even, "How might the current dominant system benefit us (even if we don't believe it is)?"
I don't favor the frame that systems "need" something, including systems "being able to change," because holding the view that systems have needs outside of the meaning and needs we assign to them belies our authorship and undermines our agency.
I'm familiar with the maintainer frame on dominant roles in H1, though I think most (maybe all?) of us play a role in maintaining current dominant systems (so instead of "people," I would offer "us" :) Even change agents who think they're advancing H2 can unknowingly feed dynamics that reinforce H1.
When I do use Three Horizons now, I tend to add the past (H0) because I find that we miss a lot about the other horizons and the conditions needed to enable movement without the truth-telling, reconciliation, healing, and learning that we gain from the past. This is especially important when working in "post-colonial" and Indigenous contexts.
Hi Russ, I like the use of H0 here. I quite often (if I have some time and space in a workshop) use a timeline that goes back at least 100 years to understand how the group’s perspectives differ on the history of the system of interest. What’s interesting is that long timelines seem to liberate people — they’ll sometime jump much further back to add things earlier on.
Incidentally, the credit on the bottom of those slides probably ought to be “Hodgson, A., and Sharpe, B. (2007). Deepening Futures with System Structure. In: van der Heijden, K., and Sharpe, B. (2007). Scenarios for Success. Chichester: John Wiley.” The trouble with the Intelligent Infrastructure Systems work is that it includes two different descriptions, in the scenarios work and in the Technology Forward Look, that Tony and Bill had ironed out by the time they wrote the chapter for the 2007 book.I can share if it helps.
Thanks Andrew. I did the earlier one since that’s where I’d first learned about it (and these are from an internal slide deck) but what you suggest makes sense.
Thanks for this Andrew. Very confirmation of what we doing at PreEmpt.Life, using strategic foresight to PreEmpt Disaster.
WHAT WE ARE BUILDING
PreEmpt Disaster — a three-layer temporal intelligence system:
Historical layer — global database of mass harm events, Harm Gap calculated retrospectively across the full record
Standing layer — current hazard estimates for known high-risk zones, updated continuously
Live layer — ALEXIS daily alerts detecting pre-event signals in real time
Incorporates and extends the existing EMS. Public good. Built on EM-DAT, GDACS, NOAA, Munich Re, Swiss Re, Sendai Framework Monitor, plus original data collection including lessons learned and after-action reports.
THE METHODOLOGICAL FOUNDATION
WHY-54 (Hazard Decision Intelligence) ratified 25 June 2026. Three components: Standing Hazard Estimate / Resilience Coefficient / Post-Event Actuals Comparison. The Harm Gap = pre-event modelled harm minus post-event actual harm. Positive = Resilience Dividend. Negative = Preparedness Debt. Counterfactual Cost Ratio and Decision Attribution Layer registered. Cross-hazard: geophysical, climate, pandemic, conflict, industrial. Master Index v4.3.
THE WHITE PAPER
Title: The Price of Knowing
Subtitle: Why the cost of ignored hazard always exceeds the cost of preparedness — and how to measure the gap before the event, not after.
Status: Section 1 drafted. Architecture complete — ten sections. Venezuela 2026 as opening case. Primary audience: emergency response organisations and organisations in affected areas.
The white paper is now the introduction to PreEmpt Disaster — not a standalone publication.
THE INTELLECTUAL SPINE — THREE AUTHORS, ONE ARGUMENT
Gill Kernick — Catastrophe and Systemic Change (2026)
The Grenfell Tower fire and dozens of other disasters examined. Her central finding: the system is perfectly designed to ensure we do not learn. Her key distinction: piecemeal change fixes a part; systemic change shifts the conditions holding the system in place. Her six diagnostic questions for systems that don't want to change are the analytical instrument for the Decision Attribution Layer.
Most importantly for the white paper: she establishes that the Grenfell deaths were foretold — in coroners' reports, in MPs' letters, in residents' emails, in cladding contractors' own communications. The word foretold is now active PreEmpt Disaster vocabulary. Use it deliberately and consistently.
Bazerman & Watkins — Predictable Surprises (Harvard Business School, 2004)
Core definition: a predictable surprise arises when leaders had all the data and insight needed to recognize the potential for, even the inevitability of, a crisis — and failed to act. Three root causes of why the foretelling is ignored: cognitive bias, organizational failure, political interest. Five reasons we are most likely to be surprised — positive illusions, status quo bias, minority interests benefiting from inaction, lack of vividness, institutional processes that suppress emerging threats.
Prevention framework: Recognition / Prioritization / Mobilization. Critical finding for PreEmpt Disaster: organizations achieve the greatest success preventing predictable surprises by adopting blanket measures across a spectrum of disasters — not addressing them one at a time. This validates cross-hazard universality directly.
The combined argument in one sentence:
Disasters are foretold. The foretelling is systematically suppressed — by cognitive bias, institutional failure, and political interest. PreEmpt Disaster is the infrastructure that makes the foretelling impossible to ignore — because it attaches a number to it before the event, not after.
That is the sentence the new session opens with.
THE LANGUAGE REGISTER
New vocabulary now active — use consistently throughout:
Foretold — the disaster was announced in advance, in the record, before it happened
Predictable surprise — Bazerman & Watkins' term, now part of PreEmpt Disaster's intellectual warrant
Harm Gap — the measurable distance between what was modelled and what occurred
Resilience Dividend — positive gap: the society built something
Preparedness Debt — negative gap: identifiable actors made identifiable choices
Counterfactual Cost Ratio — what preparedness cost versus what inaction cost
Decision Attribution Layer — the trail from harm back to decision
Systemic change — Kernick's frame: shifting the conditions holding the system in place, not patching individual failures
The problem I have with this outline is that it perpetuates many of the same controllability fallacies that lead to these collapses/disasters in the first place.
Engineering systems with more data isn't sufficient for a complex system that's teetering on a phase-shift collapse into a domain where there is no useful data.
The systems shift has to begin with not knowing and experimenting into the unknown over itemizing all knowns and engineering to them.
You're pointing at something real, and we haven't resolved it visibly enough in the framing.
The Harm Gap is not a prediction instrument. It measures the distance between what was knowable — seismic hazard, population exposure, governance quality — and what was done about it. Venezuela's fault system was not unknown. The governance capacity to respond was measurable. What failed was not knowledge. It was the decisions made, or not made, in the window between knowing and the event. That's the complicated-to-complex failure mode PreEmpt Disaster is designed to make visible and attributable.
You're right that a different failure mode exists — the genuine phase shift where prior data misleads rather than informs. We don't claim the system handles that yet. But we do have a working test model of experimenting with the unknown The live crisis layer has an explicit chaotic domain protocol whose instruction is: stabilise before you analyse. Not more data. Hold the line on values while the picture forms. That logic should be visible at the PreEmpt Disaster level, not buried in operational detail — and you've identified a real gap.
Where we'd push back: the binary between data engineering and experimenting into the unknown is false in the specific geography this system operates in. People are not dying because their governments lacked a probe-sense-respond epistemology. They are dying because knowledge that existed — often for decades — was ignored, suppressed, or never reached those with authority to act. The Harm Gap measures that distance. The Decision Attribution Layer names the decisions that produced it. That is a governance accountability instrument for the complicated domain. Not a controllability claim.
If you're working on what happens when decision support hits genuine phase-shift territory, that's a conversation we want to have.
.
I find the question, "How to work with systems that don’t want to change?" interesting, because systems are systems precisely because they have continuity. Without continuity, we have chaos.
And we shouldn't undervalue the value of continuity, of predictability, of something that we know or we're at least familiar with, versus the unknown, untested, and not-yet-experienced.
As Diane Musho Hamilton and folks working with Polyvagal Theory have noted, sameness calms our nervous systems while difference excites our nervous systems. People are encountering so much difference now, so quickly, that it's overwhelming our nervous systems. We don't have adequate practices to deal with that. And since so many of us have dysregulated, even traumatized, nervous systems, and predictability and repetition are natural salves for this, we tend to prefer the devil we know to the devil we don't, even if we literally believe it's a devil.
Whether we are for a system or against a system, whether we benefit from it or we are harmed by it, we all get something from it. At the very least, even if I resist the system, it gives my life some meaning and direction. So we all carry water for dominant systems, even if we're completely dissatisfied with them.
William James wrote in 1890 (in The Principles of Psychology), "The great thing in all education is to make our nervous system our ally instead of our enemy". That's the one element I'm not seeing in the Grenfell Model of Change, and maybe one we most need.
Hi Russ, One of my systems heuristics is Donald Schon’s principle of ‘dynamic conservatism’, which says that in response to an external shock a system will adjust by the smallest amount required to absorb the shock. When I’m explaining three horizons to people, I describe the 1st Horizon as the land of the “maintainer”, rather than the “manager” (as in much of the literature), because we need people who are interested in the integrity and stability of the current system. But it depends what the purpose of the present system is. When you read Gill Kernick’s book, or Peter Abbs’ coverage of the Grenfell Fire, it’s pretty clear that the purpose of the systems around Grenfell were either to maximise profit without regard for safety (in the case of the cladding contractors) or to minimise cost without regard for the interests of the residents (in the case of the TMO). Clearly these two sub-systems align quite strongly, and any notion of public interest is lost: the contractors were helped by regulatory collusion. In the terms of Charles Hampden-Turner’s Dilemma Resolution model, the Rock—the current system structure—gets everything it wants in the short term, but not in the long-term, precisely because it is not able to absorb the change elements that are represented by the Whirpool. Would it help you if the question were reframed as “How to work with systems that need to be able change but are unable to because they are locked into producing malign outcomes?”
I sometimes invite people to consider questions like, "What's keeping this system in place?" or even, "How might the current dominant system benefit us (even if we don't believe it is)?"
I don't favor the frame that systems "need" something, including systems "being able to change," because holding the view that systems have needs outside of the meaning and needs we assign to them belies our authorship and undermines our agency.
I'm familiar with the maintainer frame on dominant roles in H1, though I think most (maybe all?) of us play a role in maintaining current dominant systems (so instead of "people," I would offer "us" :) Even change agents who think they're advancing H2 can unknowingly feed dynamics that reinforce H1.
When I do use Three Horizons now, I tend to add the past (H0) because I find that we miss a lot about the other horizons and the conditions needed to enable movement without the truth-telling, reconciliation, healing, and learning that we gain from the past. This is especially important when working in "post-colonial" and Indigenous contexts.
I also think not just about the maintaining function in H1 but how and where people actively resist movement, often because of SCARF factors like threats to loss of status, certainty, etc. Here's an overlay of adoption of innovation onto the three horizons that I find interesting, if not useful: https://docs.google.com/presentation/d/16E6Jkowb2Jazbjki2xxZRb_sBn9Oge6n/edit?usp=sharing&ouid=112514619369648306016&rtpof=true&sd=true
Hi Russ, I like the use of H0 here. I quite often (if I have some time and space in a workshop) use a timeline that goes back at least 100 years to understand how the group’s perspectives differ on the history of the system of interest. What’s interesting is that long timelines seem to liberate people — they’ll sometime jump much further back to add things earlier on.
Incidentally, the credit on the bottom of those slides probably ought to be “Hodgson, A., and Sharpe, B. (2007). Deepening Futures with System Structure. In: van der Heijden, K., and Sharpe, B. (2007). Scenarios for Success. Chichester: John Wiley.” The trouble with the Intelligent Infrastructure Systems work is that it includes two different descriptions, in the scenarios work and in the Technology Forward Look, that Tony and Bill had ironed out by the time they wrote the chapter for the 2007 book.I can share if it helps.
Thanks Andrew. I did the earlier one since that’s where I’d first learned about it (and these are from an internal slide deck) but what you suggest makes sense.