Can models help us navigate uncertainty?

Probably not but at there end there are 10 ways to prevent assumptions becoming culture

 

Post the 2008 global financial crisis, the president of the European Central Bank, Jean-Claude Trichet, said in an opening address at the ECB Central Banking Conference Frankfurt, 18 November 2010, “As a policymaker during the crisis, I found the available models of limited help. In fact, I would go further: In the face of the crisis, we felt abandoned by conventional tools.”



The Gap

Many in executive and leadership positions have faced for a while a feeling that our models, simulations and computations are ill-suited to addressing big, messy, complicated real-world problems. There is a gap between the reality we read and measure today and the model that predicted today a short period ago.   Since we have yet to create new models and continue to utilise the same ideals, we know there is a gap between prediction and reality, and that gap is leading to poor decisions, which make us see a bigger gap. 

Decisions based on models today have short and long-term consequences. Poor decisions because of broken models lead to us missing an agreed performance metric and impact our children as they live with choices removed and unintended outcomes. 

It is accepted by some commentators that utility maximisation (the economic concept) is a poor way to navigate uncertainty.  The reason is that uncertainty means you cannot be sure you are optimising for the best thing in the short and long term. The map and directions our economic models provide cannot deliver a compass for wicked problems such as climate change and pandemics. 

A reason our models fail may be that they start with the wrong outcome preference, framing and assumption, which over time are transformed to no longer be preferences, framings and assumptions but beliefs and culture, and we cannot change them before they break — path dependency.  An assumption our utility maximisation economic model took a while back was to value the time of humans unequally (discrimination), and now technology laboured to convert the earth's unvalued natural resources. With this model, humanity ignores the consequences of previous generations' assumptions because they are now cultural, which we experience as differing views on the environment, society and climate.

Does system thinking break at earth scale?

Many of our complex models start with a decomposition ideal to break apart complexity into small discrete components that we find we can model and explore. However, even our understanding of atoms, once the most fundamental building block, actually contains secrets that we still can't fathom; such is the joy of quantum mechanics.  As we break things apart, we break connections we did not know existed, and so we cannot recombine the components and make what we started with.

Components don't combine to make the hole. I can put all your chemistry in a bucket, but it will not make you. 

If our model depends on the complete understanding of the smallest part and that the understanding is perfect and precise, then when doubt arises or the proof comes that this smallest part is not what we assumed, why do we think that the model will retain its ability to provide precise prediction?  When we study our smallest component parts in isolation, what linkages were there to other parts that we did not observe?

Our systems thinking and models depend on a form of separation, boundary, and interface that help us detail how these small components interact.  We have a duality of issue insomuch that if our understanding of the smaller part is not right and we also make an error in the dependencies about how this small part interacts, effects, affects and connects to other small parts, we end up with non-linear errors in our models.  Exponential (non-linear) errors, just like exponential growth, get big very quickly. 

There is a branch of thinking that retains the idea that we will get a large-scale model right because we should be able to model each small basic component part and the dependency between all parts, but we face a dichotomy that we don't know that if the whole can be represented by the parts, because of variables in time delays and the emergence of new recursive loops in the system, which is why our system thinking breaks at earth's scale. At scale, the earth is not a static model but a dynamic living, moving, changing and adapting system of systems.  The earth is adopting (not learning), and today's system creates a dynamic change that means tomorrow's system has different dependencies.   Our decisions impact and direct that change.

In the small backwater of behaviours, economics and social psychology, there is a discipline called anthropology, the study of human behaviours, which searches for differences and how these tiny one-off behaviours and ideas may be the future and may scale. Hidden in plain sight among all the noise and hubris is the new dynamic changes, which means that our understanding of the small parts and the dependency must adapt continually, or we must recognise that small errors at scale will give false conclusions. It is not that the emerging feedback loop existed when it started; it was created because of the dynamic nature of complex systems.   We lack right now the tools to create such new dynamic loops. This means that all our models are wrong; we just don't know by how much.

We lack right now the tools to create such new dynamic loops. This means that all our models are wrong; we just don't know by how much.

Just because all models are wrong, that does not mean they are not useful, but it does mean we need all models and cannot align to only one when we are dealing with issues at the earth scale.  This implies that 1.5 degrees are both right and wrong, as is 3.5 degrees.  Indeed the idea that any temperature is right or wrong is naive. Further, any one metric that we use as a tool for complex decision-making is going to be wrong. 

Decisions based on metrics.

The consequence of the maximisation theory that drives our model has ignored or unvalued some cost. It also maximises for a single outcome, which is never possible.  This means that the opposite side of that coin is true, that there will always be an overstatement of revenue and margin that can lead to false intentions.

We are drawn to decisions not based on costs but on the increment between revenue and costs (margin); we shortcut the complexity of the model to a single figure, ROI. Just like temperature in climate models. ROI decisions based on only costs do not exist and, therefore, as a viewpoint, it is discounted - reinforcing culture/ framing and mindset. ROI will always fail as a decision tool as there will always be undervalued costs and hurdles that mean we stretch the assumptions, but we love the idea of boling complexity down to a simple linear narrative.  

Can models navigate uncertainty, probably not, but they are the best tool we have.  Assuming they are a true compass is a dangerous game, especially if we have a culture that means we cannot challenge the model itself.  

I suppose here is the reality, system thinking will fail at the earth scale because assumptions become culture, and we don’t have any examples that culture changes before it breaks.  If the model cannot be challenged, we will have to wait until it breaks.

We don’t have any examples that culture changes before it breaks.

Leadership can stop assumptions from becoming culture

Assumptions can become part of a culture when they are repeated and reinforced over time, often without questioning or examining their accuracy or relevance.  Assumptions become a part of culture when they are widely held and perpetuated through social norms and values.

To prevent assumptions from becoming entrenched in the company culture, it's important that leadership take steps to challenge and question them, as well as to actively promote more inclusive and open-minded ways of thinking.

So we we asked ChatGPT for some specific recommendations on how to prevent assumptions becoming culture; we will leave it up to you to judge.

ChatGPT please give some specific steps that can help prevent assumptions from becoming a culture?

  1. Encourage critical thinking: Promote a culture of critical thinking and encourage people to challenge assumptions and beliefs that are taken for granted. This can help to break down the barriers that prevent people from questioning assumptions and can lead to more open-minded and inclusive thinking.

  2. Encourage and promote diversity: Encourage diversity in all forms, whether it's through hiring practices, community events, or other initiatives. This can help to break down stereotypes and promote understanding across different cultures and perspectives by exposing people to a wider range of perspectives and experiences.

  3. Foster open communication: Create a culture of open communication where people feel comfortable sharing their thoughts and ideas without fear of judgment. This can help to prevent assumptions from becoming entrenched by allowing people to openly question and discuss them.

  4. Educate everyone: Provide education and training on topics such as cultural competency, unconscious bias, and diversity and inclusion. This can help people to become more aware of their own assumptions and biases, and can promote more inclusive thinking. Share your findings and insights with others. Explain why certain assumptions are inaccurate and how they can be harmful. Encourage others to question their assumptions as well.

  5. Lead by example: It's important to lead by example. Model the behavior that you want to see in others by questioning your own assumptions and being open to new perspectives. By doing so, you can help to create a culture where assumptions are not accepted without question, and where people are encouraged to think critically and with an open mind.

  6. Emphasize transparency: Encourage open and honest language, and be transparent about the assumptions that underlie decisions and actions. This can help to build trust and promote accountability.

  7. Continuously assess and evaluate: Regularly review assumptions and cultural norms to determine whether they are still relevant and accurate. This can help to prevent assumptions from becoming outdated or misaligned with current needs and realities.

  8. Be aware of your assumptions: Start by examining your own assumptions and biases. Ask yourself where they come from and whether they are based on evidence or stereotypes. Awareness is the first step in changing behavior.

  9. Challenge assumptions: When you encounter assumptions that don't feel right, question them. Look for evidence that supports or contradicts them. If an assumption is not based on evidence, it should be discarded.

Hold people accountable: Finally, hold people accountable for their assumptions. Call out harmful assumptions when you see them, and encourage others to do the same. Make it clear that assumptions are not acceptable in your culture.