The hidden helping hand: expert reveals how AI tech is changing the industry

Chances are, while dealing with organizations during the pandemic, you noticed a huge rise in the use of conversational bots.

There’s an obvious reason for this – most physical locations were closed, so organizations had to fall back on online support – but it becomes especially interesting to note this is just a taster of an AI-centric revolution currently changing the world of finance.

“During the pandemic, banks put forward chatbots to handle the first level support like answering common questions and handle basic operations,” explains Maxime Dumas, Principal Researcher at Croesus. “Digital and self-service became pivotal for Canadian institutions, and AI played a major role in that.”

This is nothing new, however, as Dumas goes on to explain. AI has slowly been becoming an everyday part of day-to-day business life “with Siri on your iPhone, Alexa on Amazon Echo. Google Translate and other features used in many commercial products today.”

Yet there are less mainstream prospects whose commercial potential has yet to be fully explored.

“In 2018, Google introduced Google Duplex, a digital assistant that can make phone reservations for you,” Dumas recalls. “The AI can make phone calls and have spoken dialogs with real humans, without them even noticing they are talking to a robot.

"Language models like GPT-2 and GPT-3 also made huge splashes when they were introduced by OpenAI. They can perform various tasks such as composing or synthesizing texts, translating stuff, answering questions and more, without much training. They were basically trained by looking at massive samples of the Internet, and results are of such quality that humans can barely figure out it was written by a computer. For many academics, this was the first real step on the long road to Artificial General Intelligence, which is the ultimate goal in AI.”

So what are some other interesting ways in which AI is currently being applied?

“I am especially excited about the new opportunities in natural language processing (NLP),” Dumas reflects. “Unstructured information might account for more than 70–80% of all data in organizations. Text represents most of it for financial institutions. Think about financial news, statements, contracts, reports, forms, etc. Current advances in NLP allow us to extract key insights from textual data, such as named entities, sentiment analysis, and even automatically extract hand-written form data and convert it into structured information. Algorithms become more and more efficient in understanding text, both written and spoken.”

“AI can also correlate information from different sources, such as client phone calls, emails, and advisor notes, and make key recommendations to you. We have seen examples where AI could detect that a client talked about a sick child and recommend hospitals or financial strategies in the event of illness. AI can process astronomic quantities of information while the financial expert can put this information in context, building on their knowledge and experience to figure out the best way to use these insights.”