Say “AI” one more time…

The advocates for embedding artificial intelligence (AI) in every aspect of modern society risk tripping up over an inconvenient truth, being that the benefits and opportunities that AI ‘unlocks’ are not without risk. In fact, long before he rolled up his sleeves for a Twitter war (sorry, make that X war) with EFF leader Julius Malema, the world’s wealthiest South African born tech innovator, Elon Musk, was warning us about unchecked and unregulated AI.

A double-edged sword

Commenting at the recent Wall Street Journal CEO Council Summit in London, Musk rolled our phrases like “not necessary for anything we are doing” and “a double-edged sword” in reference to AI and super-intelligence respectively. He also warned of the inherent danger in developing a “genie that can grant you everything” before hypothesising that governments around the world would undoubtedly leverage AI for weapons development. Any doubters might spend an evening watching the 1968 film, 2001: A Space Odyssey. More than five decades ago, script writers could conceive of a sentient artificial general intelligence applying itself to mankind’s destruction. 

AI is top of mind among South Africa’s financial services providers (FSPs) as product providers wrestle with ways to improve call-centre productivity and fraud detection, while advice and risk practices figure out how the technology might improve intermediated distribution. In January this year, in an article titled ‘Advisers beware: AI is to robo-advice what the T-800 is to T-3000’, we considered the impact of AI on advice professionals. Our conclusion was that an AI-powered financial adviser or AI financial adviser was closer than previously imagined, and might have far wider ramifications than the robo-advice models that were in use at the time. Interrogated for the piece, Chat GPT piped up: “Financial intermediaries may indeed have more reason to be concerned about the potential impact of AI on their livelihoods, as compared to [robo-advice or advice] automation”. 

Bad actors will take AI tech to the wild, wild west

AI has been around for decades, but its application and uptake are growing exponentially thanks to the affordability and availability of data storage and super computers, among other tech advancements. Today’s main concern is that so-called bad actors create a wild, wild west in the AI development arena, perhaps posing more of a threat than the technology. “We must acknowledge that AI has the potential to transcend human nature and logic, evolving into new dimensions of intelligence; this could potentially challenge or pose a threat to the purpose and existence of humanity,” warns Wimpie van der Merwe, CEO at Global Choices. He prefers harnessing AI to contribute to humanity and find solutions to environmental and social challenges. 

His sentiments are echoed by many technology players active withing the financial services industry. Matthew Westaway, CEO and Co-founder at Voyc is a keen follower of AI and its development. “It is important that we, the humans behind AI, get on the same page to ensure AI is used with the best intentions,” he says, before pointing out that regulation might be the easiest way to achieve this. He would like to see a framework that requires companies to explain how their AI models are trained, and on what data … sort of like the detailed breakdown of ingredients you find on food products. This approach would be critical for AIs that make decisions that can materially impact an individual’s financial position. Westaway joins the growing cohort of tech-backed solutions innovators believe a pause in AI research is necessary to allow stakeholders to agree on the optimal way forward. 

Andre Symes, Group CEO at Genasys reminds us that AI is a very broad term for a whole group of differing technologies and solutions. “There are a couple of bodies that are now realising that there could be a lawlessness with the runaway speed of AI development,” he says, illustrating the phenomenon using the staggering uptake of Chat GPT, a large language model (LLM) that has many global citizens convinced that an AI can tackle QNA-type conversations on par with humans. He warns, however, that the real danger lies in the speed at which AI can carry out ‘dark’ activities such as bombarding firewalls; creating / disseminating false information; etc. 

Huge rewards for breaking the rules

Regulation or not, it is going to be a struggle to prevent bad actors, often operating with significant incentives, from breaking any rules that regulatory bodies put in place for AI application and development. “The idea is to create safe havens where people can test AI, which is great if people are willing to play by the rules,” Symes says. Risks arise due to the inability to create fool-proof barriers in a connected world, and the varying moral outlooks of different countries to the application of technology. How, for example, could you ensure rule adherence by economies as diverse as China, Germany, Iran, North Korea, Russia, the United Kingdom and the United States? 

The approach to AI varies from one financial sector to the next. Aside from its obvious utilisation in the analysis of big data at insurers and reinsurers, many insurers have found practical applications for AI-backed software in bolstering call centre compliance and productivity; identifying potential insurance fraud; and improving digital customer experiences. “We as agents, brokers, distributors, product designers and underwriters - the entire range of people who work in the insurance industry - need to start thinking about how we are going to use narrow AI in our specific spheres to focus, or improve or augment our capabilities,” Symes says. 

South Africa’s large retail banks are also wrestling with how to integrate AI into their client interactions and product offerings. An article by Prof. Yudhvir Seetharam, Head of Analytics, Insights and Research at FNB Commercial, points out that while automation and digitalisation “create many opportunities for businesses to streamline tasks and deliver benefits of ease and convenience to customers, this does not automatically translate to good customer service”. He warns against entirely removing the human element from a business that services human customers. The bank advocates for an equal measure of analytics and intuition to ensure positive customer experiences in the world of AI and other tech-back improvements. 

Humans key to quality customer experience, for now

Using Uber as a case study, the professor explains: “While Uber is first and foremost a digital platform, the quality of the user experience is ultimately a product of how well the app works and how the outcome of using that app turn out … in other words, an Uber customer could have a wonderfully slick, seamless and hassle-free experience of booking a ride on the digital platform, but if they are picked up in a dirty car driven by a sulky driver, and they feel they were overcharged, the overall experience is not going to be positive”. PS, we will let the Prof. off for not imagining the driverless Uber, which is surely not too far in the future. 

Asset manager Schroders is more interested in the firms that stand to profit from generative AI, and will therefore make useful additions to investment portfolios. A recent question and answer feedback session, titled ‘AI revolution: who is profiting from generative AI?’ shared insights from various investment specialists at the asset manager. “We already encounter AI in our daily lives, often without knowing it; think of voice or image recognition on your smartphone, or the TV shows which Netflix recommends to you,” says Jonathan McMullan, Global Sector Specialist, Public Equities. “What distinguishes generative AI from other hyped technologies of recent years is its tangible, practical nature”. 

Paddy Flood, Global Sector Specialist, Public Equities says there are multiple factors contributing to the emergence of generative AI. These include new architectures; enhanced computing power; data; and techniques for deploying AI ‘at the edge’. In layman’s terms, the final factor refers to “the AI computations being done on the device where the data is created, rather than on a distant data centre”. This development is crucial for applications like autonomous driving, where data instructions have to be acted on immediately with no latency or delay. However, the question that investors want answered is what companies stand to profit from the AI trend. 

This is not financial advice…

The usual disclaimer applies, dear reader: the following does not constitute financial advice nor does the mention of a company in this list make it a ‘buy’ at the time of going to print, or ever. Schroders’ investment experts mentioned the likes of NVIDIA, which operates at the ‘compute layer’ of AI solutions; cloud computing firms like Amazon Web Services, Microsoft Azure and Google Cloud Platform; and new businesses like Netflix and Uber as potential beneficiaries of AI. Sadly, many of the aforementioned companies have already shown significant capital and revenue growth in recent years, making them expensive. 

We close with a soundbite from Symes: “AI is not going to take people’s jobs; but people who use AI effectively. There is always going to be a bad actor using new innovation in the wrong way … unfortunately, in the AI world, the bad actor has the power to bring an entire country to its knees”. 

Writer’s thoughts:

One of the truths about AI and computers is that they do not tire… They focus tirelessly on their tasks, 24/7 365. And that means they have a hell of an edge in terms of efficiency and productivity. Are you concerned about the longer-term impact of AI on the financial services sector, and more specifically the financial advice profession? Please comment below, interact with us on Twitter at @fanews_online or email us your thoughts [email protected]

Comment on this Post

Name*

Email Address*

Comment*

 

Submit Comment