I have been reading about AI in financial services and thinking about it from a risk management perspective. What is there to be gained from this?
AI, like computers and other innovations, is a general purpose technology. One of the insights from Tim Harford , economist and journalist, about the impact of this type of innovation is that sometimes it takes time for innovations to have an impact because people do not immediately change the mindset associated with the previous technology. The main example he provides of this time lag is electricity. Following his invention of the light bulb in the late 1870s, Edison built electric power stations in 1881 in the US, and within a year electricity was available as a commodity. Yet as late as 1910, when electric motors had more to offer, manufacturing still relied on steam power. You can read about this here.
I found Tim’s explanation for this conundrum compelling. Steam-powered factories were arranged in a specific form to benefit from the single steam engine through a central drive shaft that ran through the factory. Initially, owners changed the source of power to electricity but to fully benefit from it, factories had to be rearranged according to a different logic. In addition, rearranging factories gave workers more autonomy and flexibility, and the way staff was recruited, trained and paid had to change as well. As a result, adopting electricity meant much more than simply substituting one source of power for another and the pace of adopting electricity was slow.
I think this analogy might be relevant in applying AI to financial services. AI offers a new way of powering decision making in businesses. The example of replacing steam power with electricity suggests that to get the full value of AI, financial services need to think about AI as more than enhancing or substituting for existing tools. Risk management requires a broader perspective to support decision making and achieve business objectives. I would hope that considering AI from this perspective would help financial services business to fully benefit from AI.
You can also help by submitting questions about AI from the perspective of risk management and governance. Send your questions by email (firstname.lastname@example.org) or leave a comment. I am not sure I will have answers, but who knows?
If you found this post of interest, you may want to subscribe and receive further posts by email – see box on the right hand side of the screen or click here.