By David Plummer, Group CEO, Bluefriars Group
In the first of a four-part series, David Plummer shares his thoughts on AI adoption, governance and long-term sustainability for MGAs.
The first question, put less metaphorically, addresses whether AI is sustainable in an MGA context. Much like earlier tech-bubbles, busy MGA CEOs are faced with trying to pick a corporate winner with a sustainable approach that benefits both provider and user. To me, this seems harder than it first appears.
Before I go propose solutions and mitigations to some of AI’s more obvious structural flaws, I should note that I am not against AI itself, but, rather, in favour of its sustainable deployment. Technological paradigm shifts create periods of significant disruption. I think that even a cursory look suggests we are in uncertain times. So, whilst it is the job of any MGA leader to find pragmatic solutions to market changes, any CEO or CUO who has not considered the issues addressed in this series of five articles may need to spend more time exploring AI’s risk landscape.
What is a bubble and does AI meet the test?
Just to be clear: yes, this is a bubble… …not a bubble like the famed Dutch Tulip Mania of the 1630s, although, there are market commentators saying that is possible. However, like the dot com bubble, I do not expect all AI firms to be winners in this high-stakes, high-capital commercial battle; particularly as there are competing national interests at stake in a significantly-less-stable geopolitical environment.
The Minsky-Kindleberger model of an investment bubble requires:
Displacement: A fresh approach draws investors. This is often led by paradigm-shifting technology.
Boom: Asset prices rise as capital is reallocated to this new offer. Capital appreciation attracts social / media attention and further buyers who fear missing out.
Euphoria: Considerations of intrinsic value are sidelined. Growth is asset price accelerates as both individual and institutional investors bet on the pricing trend.
Profit-taking: Canny investors perceive the intrinsic value gap and begin lock in gains. Volatility may increase, but prices may still edge higher.
Panic / Burst: A catalyst event occurs, reducing confidence. Investors shift to selling and prices fall rapidly as the bubble bursts. My personal view is that we are somewhere between stages three and four of this cycle. Other AI models are emerging, attitudes towards AI’s impact are hardening, and the environmental impact of the technology is beginning to take its toll.
Is AI Critical Infrastructure / A Choke Point?
Critical infrastructure / choke points for a business refers to assets, systems, and networks that are essential to its continued functioning. Similarly, for an economy, disruption to such systems severely impacts public safety, national security, or economic stability. When I look at the role AI systems are seeking to inhabit and, perhaps, even monopolise, I see a reliance risk for businesses. Worse I see both businesses and wider infrastructure as a target for bad actors.
Aside from this, there are the risks inherent to AI itself, which can prove as unreliable as your trickiest human asset. In fact, there are strong arguments for treating the hiring of AI with the same caution as you would treat the hiring a very-senior or c-suite employee. It is entirely possible that AI is already leaving ‘post-employment-timebombs’ inside businesses in exactly the same way a bad hire can.
From a global perspective, we are, once again, living in a world where the targeting of national infrastructure during conflict and in pursuit of dominance in nation states’ spheres of influence is being normalised. So where does that leave us? There is developing and, almost certainly, ongoing pressure to deploy AI tools to keep your MGA competing effectively with others. However, MGAs should be cautious in choosing their partners and the degree to which they rely on them. I will talk in detail about data and IP risks in later articles in this series, but, at a minimum, firms should have total clarity on data and IP use, storage, ownership, and where relevant, recovery. CEOs and CUOs should have carefully considered the sustainability of each AI relationship.
MGAs should accept that there are multiple reasons why services they come to rely on may become unavailable either temporarily or permanently and have contingencies to continue working without them. CEOs and CUOs should maintain skills in their business capable of managing this contingency. In the remainder of this series, we will continue to develop this theme by looking at how implementation and reliance on AI will begin to change the shape of your business, and what steps you should take to ensure those changes do not become obstacles to business growth and scaling.
AI has the capacity to help, but it is unlikely to be any simpler to employ than human capital in complex organisations.
Coming Next: The Church of AI | Part 2 How should I go about worship at the Church of AI, and how regularly?
Comments