Wicked problems. It’s a term that’s become ubiquitous in policy and planning circles.
The term was coined 50 years ago by two design theorists to differentiate some of the work they were doing, in comparison with “tame” problems of mathematics and chess.
It’s a term that encapsulates so many of the big, complex challenges that the world faces, including the threats of climate change and biodiversity loss.
The issue with wicked problems is that they are complex, open-ended and often intractable. One such problem is biodiversity loss.
We have already seen how a lack of understanding regarding the links between biodiversity and the delivery of ecosystem services has led to mismanagement, with negative impacts on the environment, the economy and our wellbeing.
Sustaining biodiversity and ecosystem services is challenging because ecological and socio-economic priorities can be misaligned.
Fundamental to this wicked problem is that the scale of biodiversity loss and the economic impact of ecosystem services loss are often caused by multiple threats over a long period of time (decades to centuries). Like the idea of the butterfly effect, changes that occurred decades ago can be felt today and modest increases in temperature can lead to changes in biological communities down the line and affect ecosystem functions.
Traditional ways of assessing and mitigating biodiversity loss against pollution or habitat loss don’t cut it, because they do not encapsulate this complexity and are limited to things we can see and measure. In our work at the University of Birmingham, we quantify the ‘ghost’ genetic material that has been left behind by plants, animals and bacteria to reconstruct entire communities in a single tube reaction at competitive prices.
New eras in Artificial Intelligence
But how do we make sense of changes that occurred over long time?
Seven months ago, the release of ChatGPT has seemingly heralded a new era for artificial intelligence.
The explosion of large language model AI systems like OpenAI’s ChatGPT, Microsoft’s LLM models and Google Bard has also led to increasingly polarised views about how ‘safe’ AI is. One headline asks ‘How to develop artificial super-intelligence without destroying humanity’, and Prime Minister Rishi Sunak to unveil a new £100m fund on ‘safe’ AI as well as featuring as a major talking point in a meeting with US President Joe Biden.
Why am I talking about AI?
Like for everything else with enormous potential, unregulated use of artificial intelligence can be detrimental. However, AI holds terrific promises in many fields, one of which is combining and interpreting data to tackle wicked problems like biodiversity loss.
To this end, AI can learn from past trends in biodiversity loss measure with ‘ghost’ DNA material, to predict future changes and help mitigate the impact of ecosystem service loss for future generations.
Biodiversity time machine
In our work at the University of Birmingham, we have harnessed the power of AI to look at biological data from lake sediments at an unparalleled level and create, in essence, a time machine to quantify ecosystem-level changes of biodiversity across one hundred years.. .
Combining lake sediment data alongside information about climate change and pollution using artificial intelligence (AI), we have been able to identify the specific pollutants that caused the historic loss of species observed today.
This is the first proof of concept that the time machine works to understand the past. We are now starting to build AI models that allow us to predict the likely loss of biodiversity under a business as usual or other pollution scenarios.
A tool like this can help regulators to prioritise the conservation of certain groups of species that deliver ecosystem services. It also enables us to work the regulation of those pollutants that have the biggest impact on current and future biodiversity.
While we could never fully predict the future impact of pollution on biodiversity, AI can enable us to identify the best scenarios for limiting the damage to ecosystems. Biodiversity conservation is only one example of how AI can be put to good use.
Professor Luisa Orsini MRSB, Professor of Evolutionary Systems Biology and Environmental Omics and a Alan Turing Fellow, Fellow of the Higher Education Academy
