The Canadian science-based scenario generation company Riskthinking.AI and Silo AI have together been working to produce artificial intelligence powered and science-based approaches to understand climate change risks. Riskthinking.AI is the world’s first company to generate scenarios for financial institutions that accurately capture extreme scenarios, both positive and negative. As their chosen AI partner, Silo AI has been both developing the top-tier AI models, as well as creating a manually annotated unique dataset to train and evaluate such models.
As the first risk scenario to focus on, the AI solution published today brings out insights related to one particular consequence of climate change, the rise of sea level. This global threat affects millions of people around the world in some of the most populous areas. The intelligent system bases its risk assessment and predictions on many thousands of articles. The solution analyzes vast amounts of textual data with the help of natural language processing, making documents such as newspapers and scientific journals searchable. As a result, the solution is able to extract all published predictions about the rise of sea level in a certain place at a certain time from whole publicly available literature, thus providing a forward-looking view on the uncertainty in the science of sea-level rise. This data is then input to their multifactor scenario generation model.
– With Silo AI we’ve been able to work with some of the leading scientists in natural language processing (NLP) such as Professor Filip Ginter, PhD, and Farrokh Mehryary, PhD. These methods require deep knowledge and capabilities both in terms of data collection and processing, as well as in terms of NLP and the models themselves. We’re extremely happy with the results and how the AI solution improves the scenario-based risk management for our clients. After this first edition, we aim to continue the collaboration to extend it to several other use cases too, such as evaluating the pandemic risks, earthquakes, and more, comments Dr. Ron Dembo, CEO and founder of Riskthinking.AI, a former professor at Yale University and founder and CEO of Algorithmics Inc., acquired by IBM.
– There is a lot of data out there that can help us cope with climate change, and understand the risks better. This just has to be made accessible and searchable, Dembo continues. Our goal is to create the world’s first complete risk database for climate change.
– Working with assessing macro-level risk related to climate change has been both extremely challenging and rewarding. The task presented to us required a lot of preprocessing and standardization with unstructured data, but also building tailored AI models for this unique data set and the problem at hand. Our NLP team at Silo AI is pleased with the outcome, where we managed to build a pipeline to collect and analyze vast amounts of literature and that automatically collects risks, comments Professor Filip Ginter, AI Scientist at Silo AI and leader of one of the leading NLP research groups, TurkuNLP.
About Silo AI
Silo AI is the largest AI solution and service provider in the Nordics that solves the most strenuous challenges in machine learning, computer vision and natural language processing. Silo AI aims at being a trusted partner that brings AI into product development and delivers AI-driven solutions and products. The company serves clients across several industries on four continents. Some Silo AI solutions include a tool for the Finnish airline Finnair to improve situational awareness and a tool to improve water quality at water treatment facilities for the global leading engineering company Ramboll.
Peter Sarlin CEO, Silo AI
Riskthinking.AI is a company specializing in extreme risk and, in particular, the financial risk of climate change. The company has built an algorithmic scenario generation engine, the first of its kind, which has significant benefits to the measurement of climate related financial risk.