Threat or tool? Friend or foe? Across the world, in every sector of business and walk of life the seemingly sudden emergence of generative artificial intelligence into the mainstream is prompting discussion that can be framed in those simplistic terms, writes Contributing Editor David Worsfold.
The power of generative AI to source, manipulate and re-purpose massive amounts of data at speed clearly makes it a technology that chief investment officers cannot ignore. They are already exploring its potential.
The focus so far is on operational efficiency, firmly planting AI on the ‘tool’ side of the debate.
There are the obvious tasks that AI can be almost immediately applied to where gathering and analysing large amounts of data is necessary: “The starting point for us has been using AI to realise all the benefits we were promised from big data a decade ago”, says one senior investment analyst for a major insurer.
“We can now instantly analyse once incompatible datasets, provided we have asked AI the right questions. It enables us to take what we have been doing a stage further. For instance, we have seen sentiment tracking tools used to improve our understanding of the prospects for major companies, specific asset classes and markets but now we can take that to a whole new level. We can ask it to scrutinise all statements, reports, social media, websites and broadcasts. Instantly, we get a more nuanced view.”
Reporting can be transformed using AI’s natural language skills. Asset managers are already using it when reporting to clients and insurance company boards will soon be reading reports from CIOs created by AI, potentially a liberation for those deeply skilled in numbers but less happy when it comes to words.
When will it make the leap into actively managing some or all of an insurer’s portfolio? Perhaps moving from being a tool to being a threat, at least to those in insurer investment teams.
We appear to be some way from that being a realistic prospect, although the pace of development of AI shortens timescales dramatically.
It will be tested on its ability to make ever more complex decisions but those decisions will only be as good as the questions it is posed and the data it can source. The fear, a legitimate fear, is that the source data might not be reliable or might even be deliberately misleading. The potential for threat actors, either criminal or governmental, to flood the internet with false data is real.
“We believe there will always have to be a ‘human in the loop’ to guard against AI making decisions based on flawed or false data”, says our investment analyst. The bigger and more sensitive the decisions that need to be made the more vital that human intervention.
Alongside this there will be a raft of major issues around regulation and compliance. Regulators demand transparency and clear audit trails. It seems unlikely that they will be satisfied with an answer that Chat GPT made the decisions without any independent validation of the data sources it used and the outcomes it created using that data. That, for the time being, leaves AI firmly in the mix as a tool, a friendly tool but no more than that.
There are some extreme predictions about AI sweeping away tens of thousands of highly paid jobs, especially in the financial world, but there are some more balanced insights available. Deloitte Global’s latest report, Artificial Intelligence: The next frontier for investment management firms is one of them.
It focuses on four pillars for transformation where Deloitte believes the application of AI is likely to have the most impact:
- Generating alpha: For firms seeking organic growth through outperformance, adopting alternative data sets and AI have proved to be a differentiating factor for generating additional alpha.
- Enhancing operational efficiency: Firms will continue to deploy AI and advanced automation to continuously improve the efficiency of their operations. Beyond this, firms can transform these traditional cost centres into AI-enabled “as a service” offerings.
- Improving product and content distribution: Customer experience is a new battleground and AI is helping advisors to generate more insights, customize content more effectively, and deliver it to clients with greater agility and speed.
- Managing risk: AI is a game changer for risk management. AI equips firms with the tools to bolster compliance and risk management functions, augment and automate data analysis, and anticipate and manage ambiguous events.
The report suggests that when these four pillars are augmented with AI, investment teams will be able to rapidly transform business models, operations, and internal capabilities but it concludes that “to fully benefit from AI, firms will need to carefully consider and manage the intersection between technology and talent”.
Talent – the human in the loop.