Every decision involves an element of risk. We process information on the risks associated with several decisions every day. Simpler decisions might include crossing a road, visiting a grocery store during the ongoing pandemic, or buying a new brand of shampoo. More complex decisions may include making an investment, changing jobs, or entering into a relationship. We consider several scenarios, using a wide range of known and unknown parameters, and come to a decision on what action we are going to take. Many of these processes happen unconsciously and we may not realize the amount of data we have processed to arrive at decisions to take “calculated risks”.
Decisions related to the survival and development of our society are similar but much more complex with much more data to process. The volumes of data in play truly transform into what is called big data, when the environmental and climatic context is taken into account. Meteorologist Edward Lorenz had developed the concept of the butterfly effect, according to which the flapping of a butterfly’s wings can alter the weather and even cause a tornado in a distant place. The concept tries to illustrate the complexity of short and medium term weather forecasting. Make it multiple when trying to deal with long-term climate change.
ENTER ARTIFICIAL INTELLIGENCE
Processing such large and complex datasets is only possible with the help of advanced computers and tools such as artificial intelligence and machine learning (AI/ML). Although significant work has been done in recent years on predicting climate-related risks, the science to determine the impact of these risks on people is still in its infancy and requires attention and investment.
READ ALSO | How will artificial intelligence and machine learning further advance the Ed-tech sector?
Our weather services can now warn of an impending cyclone more than a week in advance, and we have apps on our phones that can warn us of an impending local thunderstorm within a short time of a few hours. What they don’t tell us, however, is how we will be affected by these dangers.
Just as a lot of information is needed to forecast the weather, there are many things that will determine the impact that a cyclone, flood or heat wave will have on a person, neighborhood or city. The ability to assimilate and process this data will give India the much-needed edge to survive the impending next level of climate crisis. It even has the potential to show us the way to thrive despite the climate emergency, learning to contain the worsening impacts of climate change and learning to live with risks we cannot eliminate.
TO BE FAR AHEAD OF
The front lines of the current impacts of climate change are the dwellings of the economically weakest strata of society, where the capacity to absorb shocks is low. Coastal areas experiencing more frequent storms, mountainous regions experiencing an increased incidence of glacial lake outbursts, flooding, and cultivated plains experiencing recurrent abnormal rainfall or dry spells see communities devastated each year. . AI-based models are beginning to predict the hyper-local impacts of such events on citizens with a high level of accuracy.
READ ALSO | AI can’t copyright their art, US Copyright Office says human authorship is needed for patent
Hyperlocal here means that it is now possible to know and warn a vulnerable family in advance of the impending threat, and to help them understand the damage it will inflict on their home and environment, allowing them so to take the appropriate measures. An action that can save lives, and savings for life in many cases.
With rising levels of climate emergency, evident in the way unprecedented urban flooding is becoming more frequent, such tools will be critical to protecting lives and economic assets not just for the poor, but for all. the strata of society, and of the cities, states and nation. In addition to helping to avoid destruction from short-term threats, the tool will also help with long-term planning for climate risk and disaster reduction. This will range from resilient infrastructure planning to more risk-informed budgeting arrangements.
THE BENEFITS OUTWEIGH THE RISKS
As with any powerful tool, AI also has the potential for unintended and intentional misuse, which can have a negative impact. Land, infrastructure and asset risk data can be used to expose and exploit community weaknesses. Concerns have also been raised about the environmental impact of AI, given the carbon emissions related to the huge computational requirements needed to run AI-based processes for climate-related applications that are very complex. .
However, the advantages that AI brings to this space far outweigh its disadvantages. Data security capabilities and legal protocols are already in place to build safety nets around AI-based tools. The design of learning models using smarter sets of indicators, as well as better energy efficiency of data processing systems will further improve the already viable return on investment in AI systems for climate resilience.
READ ALSO | Indian astronomers detect habitable Earth-like planets using new AI technology
To harness the full potential of AI to understand and manage our future risks, India needs to urgently and significantly invest in developing AI capabilities in risk management. In addition to model development, there is a need to evolve an enabling policy environment that enables the convergence of required datasets, and to create a cadre of trained AI technologists and sensitized decision makers.
AI tools will make a dramatic emergence in the coming years anyway, as they make great business sense in offering to dramatically increase the efficiency of market-driven processes such as property development, supply chain and insurance. Their real social impact, however, will come from their deployment for development planning, risk reduction and emergency response purposes. This is where the fight against climate change can swing in our favor and be won.