Will AI takeover finance?

I conducted extensive research into this controversial debate with an unbiased view as I truly wanted to learn more about the impacts of AI on the finance industry and how it can be leveraged for business success. Also, I wanted to know whether I could still break into the industry without it being stolen by a robot.

Artificial intelligence (AI) is intelligence demonstrated by a machine that follows instructions known as algorithms, developed by an alleged species called programmers. They have made so much headway that some even have nationalities and passports. In 2017, Sophia, the lifelike robot developed by Hanson Robotics gained citizenship of Saudi Arabia!

Many say that AI will take over the finance industry. But what does “takeover” mean? If we assume it’s the extent to which finance jobs are automated, how many need to be replaced before we finally acknowledge that it’s a takeover? I am in no position to make that call. However, I can give you my thoughts on the issue, and you can decide for yourselves.

If you haven’t been living under a rock, you’ll know that the use of AI-powered systems is not new to the finance industry. To remain competitive in an ever-changing market, many financial institutions have been incorporating AI into their business as they view this as a long term cost-cutting investment, saving money on hiring people, and avoiding human errors in the process.

In the lending and risk assessment business, AI is built to learn from the past and are never biased, unlike humans. Thus, it offers a more accurate, cheaper, and quicker assessment of the eligibility of loan borrowers, leading to better-informed credit decisions.

Investment managers also use AI systems to monitor and analyse unstructured data such as news, and structured data such as spreadsheets in a fraction of the time that a human can. At the core, trading depends on the ability to accurately predict the future, and machines are perfect because they can crunch a huge amount of data light years faster than us. Not only that, but machines also don’t get tired and feel the impacts of fatigue (assuming they’re hooked onto an electrical main that gives them constant power). After staring at an excel spreadsheet for 2 hours, I honestly feel like I’ve lost vision in both my eyes and need to take a breather.

But this doesn’t mean that AI will completely replace humans in those roles. In fact, AI can eliminate tedious tasks, allowing finance professionals to spend more time doing interesting and important work. For example, assume an investment manager is generating a performance report for a client and is required to gather relevant financial information and manually complete the analysis. If this data-intensive process was automated, the fund manager could spend more time improving underperforming funds and maintaining client relationships.

Why can AI not replace all finance jobs?

The reality is that almost every occupation has partial automation potential, but generally, jobs that require extensive creativity and depth of industry knowledge, are in a much safer position than those that are repetitive and standardised.

Computers can mimic how we process information, but our creativity and how we feel, think, and respond cannot be coded in. Humans can adapt to any environment better than computers who are incapable of thinking out of the box in ways other than what they were programmed to do.

To explain why AI cannot replace all finance jobs, let’s look at the Investment Banking industry. You might think that out of everyone, the junior bankers are amongst the first to be replaced as a majority of their work involves collecting financial data, preparing slide decks, and other deal execution processes that could potentially be automated. However, this doesn’t mean that they will be replaced. If you’ve ever tried to organise a coffee chat with a banker, you’ll quickly realise that time is money for them because they are juggling many tasks and projects at once. Hence, AI can automate a lot of easy and tedious grunt work, allowing junior bankers to spend more time on work that entails deeper thought and creativity, such as financial modelling and valuation. Further, the risk that junior bankers will become obsolete due to automation is incredibly low because all of their analysis is bespoke to a particular company and structure of the transaction. Just because AI can provide more information, it doesn’t mean it will be more beneficial to a client. A computer can gather a list of 1,000 potential buyers to a seller, but a human can show 100 of the right buyers. As a result, it’s highly possible that there will be less junior bankers in the future, but there won’t be no junior bankers.

We could potentially even see banks elevate the role of junior bankers as a direct response to automation. It may become one which involves less of their ability to crunch numbers and more of their ability to be creative and propose new ideas. But whether AI will lead to the replacement of some junior bankers, reduction of working hours, or delegation of higher-value work is a tale that only time can tell.

The idea that AI can replace bankers become less and less applicable as we crawl up the food chain. Senior bankers, such as the Managing Director, are responsible for winning deals and clients, meeting companies, and developing relationships. What matters isn’t their ability to gather financial data or compile a list of potential buyers, but to get clients to sign on that dotted line, and that’s an entirely different ball game. The question is, can AI serve clients as well as humans? If no, advisory services remain somewhat protected from automation, but if yes, we have more pressing matters to address. Let me explain.

I personally don’t think financial advisory services (and any other advisory services for that matter) will ever be fully replaced by AI. I feel that there is an intrinsic trust that clients have towards their advisors. In addition to trusting their experience and perspective, clients want to feel understood, and frankly, to be able to go out and get a beer with them. If AI becomes advanced enough that it can maintain client relationships better than humans (which requires them to think and act at their own will), and clients are somehow willing to entrust a machine, there are bigger issues to worry about than whether AI will takeover finance. For example, these robots could go rogue and take control of the planet away from the human species! And if it gets to a point where the robots ARE the clients, well then…you already know it’s over.

But I don’t think it will ever come to this. Human skills are irreplaceable and there is always going to be a limitation to which clients want to leave an advisory role to a robot as the relationship is so dependent on emotions. I’d also like to revisit the importance of experiences for why clients will rarely entrust a machine. I’m no tech geek, but I find it hard to believe that you can teach a robot experiences, in the same way you and I have been developing them since birth. Computers can’t handle merger negotiations or evaluate how potential buyers will react during a bidding war. However, a senior banker with decades of not just industry experience, but life experiences (potential buyers are human too), will be able to provide more holistic advice as they understand the situation and how to react better than a robot. Technology should continue to be an enhancement to excellent service, not the answer to providing it.

Is AI controlling us or are we controlling it?

Whether people recognise the inevitability of robots or not, does not change the fact that they are coming, and simply ignoring this will lead to a chasm between those who lose their jobs to automation and those who reap the upside. But if history has shown us anything, it’s that developments have the power to change the world, but are never deterministic on their own. AI in itself is neutral, neither good nor bad, and it is up to finance professionals to embrace the automated systems and learn how to integrate them into their jobs. Because, while it may seem as though many companies are adopting new technologies, it is essential to not neglect the humans who will ultimately be driving this revolution. In the past, revolutions brought discoveries such as trains, cars, and electricity, but it didn't tell us what to do with it. We are still in control and have a choice on how they are used. This means striking a balance between viewing AI as a component in the decision-making process with humans as the ones who drive those decisions.

Verdict

We should not fall into the trap of fantasising about AI’s endless capabilities and how it will replace many of the finance jobs “today” because in reality, the state of the finance industry is a constantly evolving multi-variate equation. Technology isn’t the only variable that is increasing into the future, humans are also an ever-evolving, adaptive species. Now I’m not saying that we are going to naturally evolve into superhumans competing with AI anytime soon, but over time, as the private sector expands and economies continue to prosper, we are going to witness the establishment of new departments, business lines, and ventures in the finance industry as companies seek ways to remain competitive. There is little evidence to support the theory that mass unemployment arises from introducing robotic processes since new types of jobs always arise (digitisation may even create more jobs than it eliminates). My point is, who knows what the finance industry will look like 10, 20, 50 years from now. I don’t have a crystal ball, so I can’t accurately predict AI’s impact on the finance profession, but we can expect that repetitive, data-entry intensive jobs will be the first to be automated, followed by more complex but rules-based tasks such as tax return preparation. In the future, AI may replace some finance jobs totally, but for the most part, it will create new ones and increase the operational efficiency of others. In other words, AI will never replace every finance job and rule over the industry!

It also just came to me that the impact of AI on job security is probably the least of our worries right now as the coronavirus pandemic decimates the entire labour market. Nevertheless, I hope you gained some useful insights into this topic.

Stay safe and subscribe to be notified whenever I make a new post!