Since 1996 Daniel Erasmus, as a leading futurist and thinker, has been conducting scenario planning processes on technology, governance, education and the environment for governmental, foundation, and corporate clients. Daniel is a founder and director of the Digital Thinking Network (DTN), providing scenario thinking and process transformation as well as expert systems that bring real-time analysis of media and public discourse to forward-looking organizations. Daniel has worked with clients that include Royal Dutch Shell, Schlumberger, Sanoma, Telenor, Vodafone, KONE, Nokia, Rabobank, the city Rotterdam, and the Dutch Ministries of Foreign Affairs, Economic Affairs, and Spatial Planning. In the public sector, the DTN’s transformation work includes the groundbreaking initiative to half Rotterdam’s CO2 emissions by 2025 and to transform Rotterdam into Europe’s leading intercultural city. He lead the Economic Development Board of Rotterdam’s creation of an International Advisory Board.
Daniel authored the column The Information Society for The Financial Times Review; the column The Economy of Ideas for Intelligence and for Intermediair; and the column Hello World for PC-Review. He edited The Journal for Convergence, and the book Reflecting the Internet 1.0: Human and has authored several academic publications in the fields of multimedia, technology, risk management and the Internet. Daniel is also a judge in the $100 million McArthur Foundation Award: 100&Change.
We caught up with him to talk about the future of scenario planning.
Let’s start by going back to the history of scenario planning. I understand Arie de Geus started your journey in scenario thinking.
Yes – Arie and I were old friends. We collaborated in setting up the European Learning Network in the early 1990s and he has been a mentor to me in scenario planning or as he liked to call it scenario thinking. I recall him smiling in a rather devious way “Daniel, the problem you are bringing to me is a language creating problem and the best way to create new language is to build scenarios, and you’ll be very good at it.” My reaction was that scenario planning was just an extension of the command and control paradigm and it had little validity in a world rapidly changing. Arie saw scenarios rather differently. What he noticed was that scenario planning at its heart was a learning process where we could create future worlds/ scenarios to organize around- rather than control. It is this language creating process that makes scenarios so important, without multiple shared stories of the future organizations are left following charismatic leaders more often than not to disastrous ends.
Somewhat reluctantly I followed his advice, and at that age when a mentor tells you you’ll be good at it- you are inclined to follow. Now almost three decades later, we have created multiple billions dollars of revenue resulting from our scenario work. I never asked him if it was just a technique, but certainly thank him for that steer. He passed away early in 2020, a terrible loss.
Perhaps I should give a quick background in scenario planning. Scenario thinking, introduced in business by Shell Group Planning in the 1970s, has evolved as a powerful methodology to enable organizations to structurally anticipate, change and incorporate external uncertainty into the internal decision-making processes. Building scenarios is the process of qualitatively and quantitatively sifting, sorting and combining these possibilities into a few stories. These stories are:
- relevant – they must matter to the future of an organization
- plausible – they must describe futures that reasonably could happen
- coherent – they must have a coherent storyline
- surprising – they must challenge existing assumptions
A good scenario set consists of two to four stories that all meet the above criteria to the same degree. The stories of the scenarios should not be focussed on the developments that a client in isolation can influence, but on the developments that it cannot. The question for the scenario team is not which scenario to realize, but which innovations, human resources, IT systems, management incentives, business ideas and options to choose in order to respond to these scenarios. Scenarios create a language for talking about the future.
Scenarios are not predictions, nor is it crucial that the scenarios accurately reflect the future. What is important is that the scenarios are alive in the hearts and minds of the managers and leaders of the organization. The scenario process is a tool to learn from possible futures in order to make more informed decisions today. The strategic conversation that results from the scenario process and the options that each of the public and private organizations generate based on the scenarios, are their true value, not their predictability.
Shell still does great work, their recent set was fascinating, but scenario thinking has also moved forward from Arie’s days of Shell and the fantastic work done in the 90’s by Peter Schwartz, Napier Collins, Steward Brand and Kees van der Heijden at the GBN.
How is that?
In terms of scope, scale and technology. When we make scenarios, it is often 1000-2000 pages of work, a network of some of the most remarkable people on earth and the ability to task our AI platform with the problem. We are reading billions of articles and clusters of machines facilitate these insights.
Today, at the Digital Thinking Network we talk about heuristic strategy. Scenario thinking is a part, an important part, but a part of a holistic commercial, cultural, and cognitive process. What story of tomorrow does your organization want to create, or become part of? We need to get beyond the illusion of control.
At the DTN we developed strategic heuristics to constantly update what used to be called strategy based on an interplay of learning, sense-making, situational awareness, language creation within the context of a changing environment. John Seely Brown calls it “small moves smartly made”, I prefer the term heuristics, because I’m old school, and like the root “I find, discover”.
Wait, can you explain what the Digital Thinking Network is?
The Digital Thinking Network (DTN) combines a worldwide network of experts, scenario thinking and AI/big data computing tools to create a path for organizations to navigate through the “whitewater world” – as JSB calls it.
The worldwide network of experts includes former prime ministers, corporate CEOs, academics, artists, former generals, former hackers, cultural leaders and pioneers in new media technologies. The DTN regularly conducts interviews with these leaders to capture their perspectives on a changing landscape and gain diverse insights into why things happen and how they connect. These are typically people behind the news, occasionally you would see their name appearing but more often than not you would see their ideas progress through networks and news. I believe we have done just under a thousand of interviews with this network of remarkable people. These are terribly long and deep interviews that take hours, I recall a 5+ hour interview with Alan Kay, where both sides did not want to stop and he was left running to catch his plane. Whether we are assembling an international advisory board for a major city or challenging a prevailing mindset, our network is the fountain head of thinking in the DTN to give a broad spectrum of experiences and views to induce fresh thinking and breakthrough ideas.
Over the last 20 years The DTN has facilitated more than a 100 scenario sets exploring futures with global thought leaders from inventors of the Internet Vint Cerf & Bob Kahn; entrepreneurial pioneer of mobile payments Takeshi Natsuno or founder of Rakuten Shinnosuke Honjo; to Nobel Laureates George Akerlof and James Mirrlees; to author of Limits to Growth Jorgen Randers; climate change discover scientist Jim Hansen; to CEO of SWIFT Gottfried Leibrand, CEO of TCS Natarajan Chandrasekaran; to one of the sushi masters of the Emperor of Japan Eiji Sato-Oyakata and HRH Prince Carlos de Bourbon de Parne, etc.
We combine the insights gleaned from these interviews to create an Early Learning System which builds on scenario thinking with planetary scale natural language processing built on AI/ Big Data analytics to algorithmically gain edge insights from very, very large corpi, think billions of documents, of unstructured data.
Why? Because learning better is the only sustainable competitive advantage of the 21st Century as Arie de Geus often stated. We are going to have to learn a lot better, fast.
How does a business derive value from being part of the Digital Thinking Network?
Either though breakthrough strategies, preparing for external shocks, acquisitions or organizational cultural change. It shifts the organization from a culture of fearing its external environment and uncertainty to using these insights for breakout success.
We anticipated the global financial crisis (in early 2006) for The Netherlands’ largest bank, the scenarios are written up pre-crisis, and the bank made a profit in 2008, during the GFC. They made more than $2.5B profit, their largest profit in their 150 year history in 2008, while the entire industry was collapsing around them. Much of this was written up in a book, which they kindly allowed to become public, and the work additionally anticipated the global fintech revolution more than a decade ago: The future of ICT in financial services: The Rabobank ICT Scenarios. We printed 10000 copies of the book, but most of the copies was circulated in private.
Strategic acquisitions resulting from DTN scenarios have created $350 million in profit in 3 years- from a € 390 million sale of a strategic acquisition bought for € 37 million. Again the Rabobank bought Alex a small firm, the purchase was informed by our scenarios that it was a robust strategic option, something that the board supported despite significant cost cutting in the bank at the time.
Scenarios together with our AI analytics platform anticipated the 2014 Oil Price Collapse in 2012 for a large Oil Field Service company informing a $50 B merger in the Oil Field Sector and additional $3.5B in profit. We were discussing the nature and triggers for an oil price collapse two years before it occurred. In our AI system we saw the sentiment change, a few months before the events. Without the scenario background I’m not sure we would have understood what the shift in sentiment meant, the combination was really a breakthrough.
For governmental clients DTN scenarios anticipated the $ 70 oil price when it was $23 (2003), unfortunately the client never managed to secure the long term energy contracts which would have resulted in extremely large sums. Reality is always complex and scenarios are not a guarantee of success- nothing is. It is the interplay between the new vistas that the scenarios open, the decision making process, and the ability to track the scenarios at scale that in our experience have delivered the breakthrough successes.
As you can see there is a real commercial story behind these scenario strategies. There must be, because it is a lot of work, and really should create commercial value for the clients. I always say “you are as good as your advice taken, not as good as your advice given.” The real proof is, do organizations make different decisions based on these learning processes, and that is an aesthetic Arie and later JSB always instilled in me.
You also redesigned the Shell breakthrough innovation process called GameChanger. What was that about and how do you help businesses rethink innovation now?
Shell’s GameChanger when it was created in the 1990’s by Gary Hamel and the Shell team, has been a fantastic success in breakthrough innovation. By the 2010s its head Russ Conser and Hans Haringa realized that the process goals needed a redesign for the 21st Century. We were honored to be asked to redesign the process based on the new challenges of low carbon energy, globalization and the awesome capabilities that AI would create. Let me describe 3 thinking mistakes often made in thinking about early stage innovation.
Early stage innovation is people centric not idea centric:
This is incidentally the mistake that some of the European funded innovation programs make- they approach it in an idea centric, naive planing way. One of the things that people get wrong in radical innovation is they try to assess the idea based on a description of that idea. If it’s a suitably different idea, one that challenges existing assumptions, you just don’t have the language for it yet. You don’t even understand the parameters of success. You don’t understand what it would look like if it works. In fact, you have an idea and you think it might be useful for one thing, but it turns out to be useful for something completely different.
A key part of the radical innovation process therefore is to bring people into conversation on that idea — a conversation with something like a prototype to anchor it. I coined the term diagetic prototypes to describe this process.
Secondly, there is no forward, there is only outward.
How do you navigate a forest? What does progress look like when you are not sure which direction is forward? Early stage innovation is akin to navigating in a rainforest- in the dark. You have a notion of what you are trying to achieve but the maps describe elevation not territory. Simple notions of planning do not survive the complexity of these edge environments. A promising technology can meet with an insurmountable obstacle, after which it gets appropriated for another purpose only to connect to an adjacent enabling technology that makes it fit for the original intent. This circuitous route is impossible to plan for, and one is constantly operating with partial information. This is not a free for all, but purposeful exploration, reflection, “fastest path to failure” principles, with a deep understudying the aesthetics of success. There are no maps here.
Another thing done wrong in radical innovation is people try to assess an idea based on a description of that idea. It will be assessed through play. Russ Conser calls this the “speak to the ghost” principle, quoting Dickens: “An idea, like a ghost, must be spoken to a little before it will explain itself.” The grammar of the ghost is play.
Not planning – playing!
GameChanger navigates the forest through play. Not as a desk exercise or a spreadsheet simulation, but as in the complex interaction which we call P–Cubed. People Playing with ideas, through Prototypes. In learning by playing, the proponents and GameChangers map greater and greater parts of the forest, moving outward. This is done through small projects, funded as little options, in each case de–risking the next step. Similar to Silicon Valley’s Angel, from series A, to C, funding is done in parts. The initial step however in the GameChanger process is unique: it is the fastest path to failure.
Introducing low-threshold activities at an early stage, activities that create interaction between the idea and something in the external environment (a customer, a physical object, something), transforms it into play. When you play with an idea, the idea forms and your understanding of that idea forms. People propose an idea and then they’ll study it forever as though thinking about an innovative idea will lead to an understanding of whether it’s good or not.
The people principle is uncannily hard to practice, and almost anti-bureaucratic. Machines do not (yet) identify these type of people well. The identification of people, cultivation of networks, building of trust, coaching and mentoring through the development play of people, ideas, and “prototypes,” and finally landing the project within an often hostile audience in Shell is what makes GameChanging so hard.
Wow. Can you give us a concrete example?
What was fascinating about the redesign is that we tested a planning team against our AI using a set of themes called the Future Energy Technologies. These future energy technologies represent investment opportunities or areas to track in the outside world. When thinking about electric mobility, in 2011, the human team was thinking about plug-in hybrids, lithium-ion batteries, charging stations, range anxiety. Today we would add Tesla, the amount of lithium in Bolivia and transition costs to the list. But, unless one spends an extraordinary amount of time looking for the edge, you will only see the centre. What the human augmented AI platform noticed was at the edge, a cluster set on Electric Scooters. Even today, a decade later, the discussion about electric mobility barely considers electric scooters as important.
The performance characteristics of the batteries with limited energy density and limited range are better suited for short light city trips- exactly the type of commute we see in Asian megacities. The youthful demographics of these cities implicate single person rather than multi-person journeys as prevalent in the “old West.” Sourcing: many of the batteries are being manufactured in close proximity to the cities. Congestion: as scooters command a fraction of the available roadway space, we can expect government policies to support them. Air quality: the brooding social conflict created by the air pollution burden in the Asian megacities imply governmental support for a shift to support electric scooters as an affordable, local approach to make this transition. Lastly disruptive innovation theory as exposed by Clayton Christensen imply that inferior niche products will move up the value chain, exploiting more and more valuable niches over time, to eventually dominate incumbents. Perhaps the electric cars of the future will not be Teslas or BMWs but some of those small scooter companies that dot China. Today electric scooters outsell electric cars by several orders of magnitude. We expected 100s of millions of electric scooters on the road before 2020.
This case gives a view on the shape of innovation to come, despite some of the smartest people on earth working on this, they had missed this key insight that transforms the replacement curve, the steel required, etc. The entire domain is different because of an edge insight, spotted by our AI. Radical innovation in the absence of these tools, in the age of a world in transition, will operate largely blind.
Can you give an example of the Erasmus.AI platform in use in innovation now?
To pick a recent example for one of largest technology companies in the world we identified we identified 153 specific technology breakthroughs in 7 domains. These domains we large scale global mega-trends such as Climate Change, Energy Transition, Urbanisation, Digitisation, Food and Security, Health, etc. The team performed a whole systems analysis of each of these domains that explored how system attributes relate to each other, create dynamic responses, model the experience curves around key enabling technologies, etc. We modelled down to the detail that predict the Co2 reduction of a dollar spent on a solar panels versus the same dollar spent on clean transportation, or transition costs for regenerative agriculture. This was done both as a cognitive social process and an AI exploration in a series of sessions. Imagine 20 meters of detailed system analysis using a process that I invented in the 90’s Natural Language Systems Modelling, then deepening it in terms of key quantitative data points- I believe we had a few hundred and built quantitative models on this hard data. These different scenarios were simulated, and mapping that into our Erasmus.AI system. Throughout the process the AI platform supported this exploration, to map edges through interactive visualisations, novel ways of looking, key forces to track, inflection points, etc. This level of analysis would not be possible without our AI in combination of with deep human analysis. It is a transformative platform, but more than that it is a transformative relationship as these modes of thinking build on each other.
At the end of each of these mega-trends we identified 153 highly specific key breakthroughs or platform enabling technologies that can “change the game” in each of these domains. Think of specific technology interventions such as reducing the cost of fuel cells by 90% or radically different approaches at a systems level by redesigning desalinisation systems and technologies, MRI’s that cost a dollar, etc. We are tracking these emerging approaches in an interactive dashboard with a 100 000s dimensional model underneath it where one can navigate from high level overview showing veracity of innovation in these domains down to the level of a single article, patent or press-release. I’m not aware anything of this complexity has ever been built, and at the same time is completely intuitive- one can learn the system in a few minutes. It is the one big picture view to see how things could change.
This gives an indication of how powerful these language generating systems and processes, built on top of these scenario insights have become, and how impossible it will be to get anything done in their absence.
What have you learned about innovation and the organisation?
If we look back, the managerial response in the 1990’s and early 2000’s to place incubators outside their parent organisations was inherently a defeatist notion: an inability, or unwillingness, to respond to the scope and scale of the changes that the early Internet presented. Put deep innovation in a small organisation separate from the mothership would “not be encumbered by the rules and regimentation of large transnational organisations, this small band of merry pranksters, could operate freely, at Internet time.” However, most incubators failed because of this separation – even good ideas could not be turned into reality as they struggled to be integrated back into the parent organisations. The edge was forced to move close to the core. Guess what? This is still a problem today. We need to bring the centre closer to the edge. This is unbelievably hard.
How should businesses organise themselves to be more innovative?
In Shell’s GameChanger the view was segmentation not separation, and it need immense skill in translation, playful engagement, and institutional memory. Once the GameChangers slipped into the internal innovation poster sessions at Shell and stuck a sticker on GameChanger Inside as a play on Intel Inside on each project which was funded originally by GameChanger. To everybody’s surprise 80% of the projects had gone through GameChanger, something that even the people in the projects had forgotten. It is that playful engagement with the core, that keeps the edge valuable in the minds of the managers.
Ashby’s Law states that “only variety can absorb variety.” Systems are viable only if the variety of internal states match or exceed the variety of external states.
In the case of an edge innovation process, there should be a deep and purposeful avoidance of typical team members. The managerial response to conjure a typical profile and hire accordingly is avoided. The innovation team is formed by a group of people with as much differences between them as possible. In GameChanger surprisingly little effort is spent on ensuring that the team work synergistically together- it is not needed. What is needed is a lot of different people organising themselves in different ways around the innovation vision, and then going about in their own natural way.
We identified 7 heuristics of GameChanging innovation and then amplified that through 3 deep responses that the team could make ensure the practice renews and stays relevant in the 21st Century. One of them was to augment human intelligence through machines in a very different way than current AI rhetoric would suggest.
Prof. Dave Sinclair, the force of nature, that heads Harvard’s life extension lab, slams his fist not the table after using Erasmus.AI. “Daniel, in seconds you showed me what it has taken me 20 years to learn in this field”. When one can accelerate learning to see in seconds what the best people in their field took decades to learn, it gives a sense of what is to come. This is the space we want to enable. This is a real role for humans in the future of organisation, but it has to be in right relation to the unbelievable power of AI. Not to replace people but to enable better “magic circles” as Huizinga would argue.
Thanks so much, Daniel.
INTERVIEW BY CHRISTIAN SARKAR