How can we build a culture that will allow for adoption of new technologies?
Martin Thorn, Senior Manager – Decisioning, Analytics & AI, CYBG
Martin’s experience working for CYBG (the established brands of Clydesdale Bank and Yorkshire Bank) makes him an expert speaker, representing a company that offers a full range of financial products and services to help people and businesses grow. Ahead of speaking at TechNOVA: AI in Financial Services, Martin has shared with us some key trends he has noted since working in AI at CYBG, including the hurdles from moving from experimentation to production, building a culture to enable adoption of new technologies, and the issues AI terminology can have when being implemented across a business.
There are thousands of opportunities to utilise AI but moving new technologies from the experimental stage to production can be difficult. What are the main hurdles here and how can they be overcome?
I think the main challenge to the integration of AI systems is the hype that surrounds this technology. Everyone has read a scary story about AI and as such even fairly basic ML can generate real concern across the business. Unfortunately I haven’t found any silver bullets to this problem, but keeping experiments small and manageable and getting people comfortable with the smaller uses of new technologies is vital before you progress to more complicated items.
“The main challenge to the integration of AI systems is the hype that surrounds this technology.”
How can we improve communication within a business to create better cooperation between teams?
New technologies can both help and hinder progress as people tend to turn into keyboard warriors, passing tickets to each other rather than chatting. We have had success with creating temporary cross-functional teams, co-locating them, and then disbanding them after the project has been completed. It feels like a lot of upheaval, but the benefits of across-the-desk chat is invaluable.
“The benefits of across-the-desk chat is invaluable.”
People can be a challenge to a tech revolution. How can we build a culture that will allow for adoption of new technologies?
In my experience the main issue is that people are scared of new technologies that they don’t trust or understand. For me, the key to building a culture accepting of these technologies, is to burst the hype bubble and demystify the technology by relating it to technology or processes that people do understand. If we leave our egos at the door and admit that AI and ML aren’t quite as revolutionary as we’ve been telling people then I think people will be much less fearful and more comfortable. Moreover, it is necessary for us as technology professionals to make good on the promises surrounding new technologies – ensure that the goals we set are realistic and that we enable the benefits to be realised. Every large IT project “failure” is more ammunition for the sceptics.
“If we leave our egos at the door and admit that AI and ML aren’t quite as revolutionary as we’ve been telling people then I think people will be much less fearful and more comfortable.”
In what ways can AI terminology prevent progress in AI and impact its implementation across a business?
The very hype that generates business interest is the things that holds us back. Machine learning implies a culture where the machines get a “leg up” from the humans and then they are off learning and adapting as they go. We all know where that ends. The robots take over the world and decide humans are parasites. Most companies are miles away from the stories people read in the press about computers developing their own language, but the stakeholders don’t know that. So as an industry we need to demystify, dehype and make everything easier to understand. The more we hype the fact that the machines are doing things that we couldn’t possibly comprehend, the more people will get spooked and the more people will want to control and slow the roll out. No-one wants to start something that we can’t stop or can’t control – so the key in the terminology needs to be the concepts of checks, balances, supervised learning and development. Once you take AI process back towards terminology and concepts that people are already comfortable with then people should be more comfortable with implementation.
“As an industry we need to demystify, de-hype and make everything easier to understand.”
Martin joined Clydesdale and Yorkshire Bank Group in 2017 to grow the data science and AI capabilities of the bank. His team is responsible for revolutionising how the bank uses AI and data to create true personalisation and value for customers.
Prior to CYBG, Martin spent 10 years at Sky in various contact centre technology roles – managing large teams and developing Sky’s data programmes. With a background in marketing and analytics, he has a wealth of experience understanding over-hyped technology – as well as employing them within an organisation.
Martin will be speaking at TechNOVA: AI in Financial Services, taking place in London on 27 March 2019.
From banks to start-ups and insurers to investors, TechNOVA: AI in Financial Services will gather together the greatest minds to expand understandings of AI, and build practical knowledge on how to develop real, implementable systems. Don’t miss this fantastic opportunity to explore the everyday implications of AI, discover implementation how-to’s and learn how to adapt your AI strategy to the changing customer.