AI Summit Highlights Benefits for Insurance and Finance sectors
The world-leading AI event, World Summit AI Americas, returned to Montréal in early May. A host of inspiring speakers demonstrated how different vertical sectors are applying AI for competitive advantage. So, what can insurance and financial organisations learn from the event?
Creating a healthy data culture
AI has massive implications for the insurance and financial sectors, with the potential to deepen customer insight, boost customer experience, detect fraud, automate claims processing, and improve decision-making.
Many insurance organizations, for example, have large amounts of data, but they are not structured and prepared adequately to exploit them in projects. Also, not all insurers’ artificial intelligence departments have yet reached a level of maturity that allows them to effectively identify potential use cases with the data they possess.
Since 5-10% of data drives 90% or more of business value[1], it is far more important to focus on quality rather than quantity as much of that data is irrelevant and creates noise. This can lead to bias which will drive AI models to unrepresentative conclusions. Making sure, then, that data is correctly selected, structured, prioritised, cleaned and error-free will lead to more powerful AI-driven insights.
One of Google’s presentations highlighted the difficulty of using AI to analyse data. While the audience were able to agree on the classification of domestic cats and dogs, a picture of a tiger created disagreement as to whether it fitted into the category of cat or dog. Not everyone knew that tigers are a member of the cat family. This kind of subjective decision-making can also create bias and needs to be considered when building AI models.
A lack of understanding between data scientists, AI professionals, R&D and management can also create issues, as these individuals will have different focuses and different views of the same problem. It is essential to make sure that the objectives of an AI project are clearly understood and the type of model and kinds of metrics are agreed and defined by all stakeholders.
Bridging the gap between the academic and commercial worlds
Another topic that was discussed at the event was the need to bridge the gap between the focuses of the academic world – such as increasing the precision of AI models by fine degrees and publishing research papers – and the performance-related focus of the commercial world. The ground-breaking and incredibly complex AI models that universities may produce to solve a problem, for example, are unlikely to be easily deployed and integrated into commercial environments.
At the event, a speaker from Intact Financial Corporation spoke about how the organisation is working in partnership with a university on business-related issues by defining common goals between the two organisations, and how this also helps prepare postgraduates for the world of work.
Similarly, GFT is in discussions with academic organisations to provide postgraduate students with the opportunity to work in the commercial field in carefully selected areas. This allows academics to further explore the practical applications of their research and provides GFT with greater access to highly qualified experts and the latest advances in research.
If you’d like to explore how AI could boost your organisation’s competitiveness, contact us now on insurance.coe@gft.com.
[1] https://www.accenture.com/us-en/insights/artificial-intelligence/scaling-ai-giving-data-its-due
