Event Recap: iCAN Innovate Conference - "AI, Skills, and the Future of Work in Insurance”


Last Friday, 22 November 2024, iCAN welcomed almost 300 members to the inaugural iCAN Innovate Conference.

Hosted by our sponsors Arch Insurance at their office in Fenchurch Avenue, London, and Ki Insurance, the conference brought together professionals from across the insurance industry to explore the role of AI, emerging skillsets and the evolving workplace.

With insightful keynote speeches, panel discussions and networking opportunities, the day sparked meaningful conversations about the challenges and opportunities shaping the future of work in insurance.

Visit our Events page to register for future iCAN events. iCAN members receive invites to all of our Events, so register below if you are not already a member:

Following introductions from iCAN Co-Chair Kishan Mangat and Arch Insurance International Chief Operating Officer Ketan Motwani, our compere for the day Dr Rashada Harry (AWS and Founder of Your Future, Your Ambition (YFYA)) took us through the agenda and introduced the first expert speaker of the day, WPP’s Chief AI Officer and Satalia CEO Daniel Hulme.

Daniel’s talk – ‘Rethinking AI and its impact on business and humanity’ – started with some useful definitions of AI and explained some of the limitations of LLMs (Large Language Models), before going into more detail on the ‘Six Applications of AI’:

  1. Task Automation - Macros, RPA, chatbots, object recognition

  2. Content Generation - Images, video, text, music

  3. Human Representation - Deepfakes, voice, personas

  4. Insight Extraction - Machine learning, data-science, analytics

  5. Complex Decisions - Optimisation, decision trees, expert systems

  6. Human Augmentation - Exoskeletons, avatars, cybernetics

After a quick run-through of some AI use cases, particularly in HR and regarding organisational network analysis, Daniel then explained the slightly more scary idea of an ‘AI singularity’ – the moment when AI becomes more intelligent than humans.

According to Daniel, there are six AI singularities:

  1. Political - When we no longer know what is true

  2. Environmental - When we create uncontrollable ecological collapse

  3. Social - When we cure death

  4. Technological - When we create a superintelligence

  5. Legal - When surveillance becomes ubiquitous

  6. Economic - When we automate the majority of paid labour

Daniel’s talk resonated with the audience, who were all keen to ask questions and learn more about what AI might mean for the future of work.

Connect with Daniel Hulme on LinkedIn.

Next up on the conference stage was Jonathan Mascie-Taylor, Managing Director of Render, with his insightful talk: ‘GenAI in Insurance: Revolutionary innovation or passing trend?’

Jonathan began by explaining what is actually meant by the term ‘Gen AI’ and how it ‘generates’ new content using machine learning.

He then took the audience through the three classes of Generative AI models:

Individual force multiplier

  • Supporting individual team members to do their job quicker

  • Examples include helping underwriters / data scientists / developers understand terminology and develop code rapidly

  • Off the shelf models and UIs (ChatGPT, Google Gemini, etc.)

Augmented workers

  • Tailored solutions to specific workflow tasks

  • Examples include data extraction from documents and legal text analysis

Customer facing

  • Dedicated external facing customer chatbots

  • Often a bespoke build with large, dedicated teams needed to setup and manage

After a live demo to show how easy it is to make a fun dinosaur game using AI, Jonathan then took the audience through some potential insurance industry use cases for AI. These included:

  • Supporting compliance and legal teams. For example, loading the FCA handbook into a chatbot so compliance officers can quickly check, reference and understand relevant terms.

  • Claims management. An example could be using LLMs to quickly summarise the available options available to settle a claim e.g. highlighting differences in garage quotes for car repairs, to help select the most appropriate way forward.

  • Feature generation and extractions. This could involve using LLMs to help ‘structure’ a large amount of input emails, word documents, websites, etc. into commercial underwriting and other risk models.

Jonathan stressed that the key for using AI effectively in the insurance sector is to train tools like ChatGPT to produce the insights you need – such as the chatbot example above for compliance teams. This involves embracing the technology and using it in specific ways to improve efficiencies and decision-making processes in your organisation.

Connect with Jonathan Mascie-Taylor on LinkedIn.

Next up, Mel Zhang (Head of Algorithm Pricing) and Zhe Feng (Head of Algorithm Engineering) from Ki Insurance took us through how AI is used to make their data-led business more efficient.

Mel started off by detailing Ki’s journey so far, from their launch in 2020 to where they are now with their Ki Digital Services platform which promises to mark a step-change in how business is transacted in Lloyd’s.

Mel then opened up the floor for a quick Q+A session, with a lot of questions from the audience about the specifics of how Ki’s algorithm-driven quoting system works and where humans come into the process.

Zhe Feng then took over to take us through Ki’s journey with AI in more detail – starting with a breakdown of Ki’s algorithm today and what it is capable of in terms of granular risk assessment and portfolio management.

Zhe then presented an interesting timeline detailing when Ki initially started experimenting with AI – around 6 months after the release of ChatGPT – and showing all the steps between then and where they are now with their ‘initial industrialisation’ of Generative AI and their own in-house ‘ChatGPKi’ product.

Zhe finished things off by showing a great example of how ChatGPKi can be used for slip analysis to find relevant data and wordings and ultimately increase operating efficiency.

Connect with Mel Zhang on LinkedIn.

Connect with Zhe Feng on LinkedIn.

The final talk of the day came in the form of a workshop delivered by Jonas Haefele, AI Researcher & Somatic Coach at Slow Works.

Jonas’ thoughtful and practical workshop introduced the audience to new ways of thinking about AI and how to maximise their use of tools to “intentionally use AI”.

The audience were asked two questions to help them get a better understanding of how AI could be used to benefit them:

  1. Which aspects of your role would you prefer to delegate?

  2. Which aspects of your role would you like to have specialist help with?

The answer to both of these questions is where you could start using AI to really help your productivity, efficiency and development.

After a valuable section on the importance of using AI safely and responsibly, Jonas then delivered a great lesson in prompt engineering. Starting off with the important advice of “new job, new chat” to make sure your results from ChatGPT are not polluted by any unrelated previous requests, Jonas demonstrated several prompts you can use to get more focused results from the tool.

For example, starting your prompt with “Act as a Lloyd's underwriter specialising in [risk]...” immediately allows ChatGPT to be in ‘insurance mode’.

Another useful tip is to include instructions in the prompt for ChatGPT to ask you further questions for context e.g. “Your task is to [write a memo]. Ask me at least 3 questions, one at a time, before you write the memo.”

Following this exercise, Jonas stressed the importance of critical thinking when using AI to make sure you are controlling the results you get.

Connect with Jonas Haefele on LinkedIn.

The final session of the day was a panel discussion moderated by Kishan Mangat and featuring Ed Gaze (CEO of Innovative Risk Labs), Maddy Robson (Head of Strategy, Vision and Business Change at Allianz), Claire Tunley (CEO of the Financial Services Skills Commission) and our compere for the day Dr Rashada Harry.

Some of the topics discussed included how the insurance industry is still combining new technologies with old mindsets, and why it is important for all of us to get out of our comfort zone and experiment with AI tools to find what works and what doesn’t.

Perhaps the biggest takeaway from the panel discussion was regarding the topic of what future skills will be important as AI becomes more prevalent in the workplace. The panel were unanimous in their opinion that the ‘soft skills’ that humans can bring into the workplace will be much more valuable in the future, such as communication skills, empathy and critical thinking – as AI cannot easily replicate them.

Connect with Kishan Mangat on LinkedIn.

Connect with Ed Gaze on LinkedIn.

Connect with Maddy Robson on LinkedIn.

Connect with Claire Tunley on LinkedIn.

Connect with Dr Rashada Harry on LinkedIn.

About iCAN

The Insurance Cultural Awareness Network (iCAN) is a leading organisation committed to promoting diversity and inclusion in the insurance sector. Established in 2017, iCAN has been instrumental in spotlighting cultural awareness, fostering talent, and championing diversity within the industry.

Visit our Events page to register for future iCAN events. iCAN members receive invites to all of our Events, so register below if you are not already a member:

Previous
Previous

Happy Hanukkah!

Next
Next

MatBlas confirmed as iCAN Bronze Sponsor!