One problem with AI is its black-box nature when solving problems. It will not be specific about the indicators that lead to a solution. Generally speaking, Elin Hauge, an independent AI and business strategist, believes we should first look at the problem we are trying to solve, and then look only at the toolkit that includes AI, among many other tools. Solving a problem is sometimes “just getting the flow of data between two different systems to work properly.”
It won’t tell you why it’s weird, but it will tell you that this data point is weird
She observed that many executives’ decision to use AI is “largely” driven by consulting firms. “If McKinsey says it, it must be true, right? » Hauge commented that understanding data dynamics requires a lot of skill in an industry, its data and its value chain. Yet such context helps decide whether mathematics based on stochastic modeling should be used or not.
AI in the financial sector
While discussing the suitability of using AI to predict certain market indicators such as stocks or interest rates, AI could be more appropriately used to detect outliers, Hauge suggested during an interview on November 27, 2024, following his presentation at a European Investment Bank conference. “It won’t tell you why it’s weird, but it will tell you that this data point is weird.”
She noted that technical experts at financial institutions “kind of use” large language models when writing internal code development. “It has nothing to do with trade.”
Can insurers and banks benefit from AI?
She noted that among insurers, machine learning or “good old stochastic modeling” is used for “micro-pricing of individual risk level.” Elsewhere, claims fraud detection is another sector benefiting from AI. Hauge explained that machine learning helps identify “patterns of behavior in known fraud cases,” also called “pattern recognition.” In the financial sector, the credit card industry views the issue from the opposite angle, that is, “a departure from a typical model.” An alarm would go off if your card was used in the Bahamas, for example.
She believes that “good old-fashioned statistics works really well when you have a predefined data set” to perform regression analysis on. On the other hand, Hauge asserted that when faced with “a larger data set and you don’t really know which data is more important to you and which is less important in pricing, then the Machine learning is one way I go about this.
A surprising statement for your correspondent given that there are statistical methods for selecting relevant variables. Hauge confirmed that the approach or “the math behind it is the same,” but doing it with AI is a faster way to achieve your goal. However, there is a risk of losing information on material indicators along the way.
Identify relevant indicators and their impact. All is not lost.
“There is a way to solve this problem one way or another, using heat maps in segments of your neural network, and AI is one of the similar subfields of technology currently being developed and studied,” Hauger said. She said neural networks with “many layers are called deep learning (systems),” which is the method used for most of these models.
She noted that regulations require data transparency and the use of algorithms for “higher risk applications.” Therefore, heatmaps attempt to assign weights to neural network nodes, “a sort of black box.” However, she admitted that it is “not mature enough… but it is an area in progress”.
Humans are still needed to support AI
Hauge commented that “any of these models” provide stochastic predictions. She suggested that a result with a 95% probability level could mean that “you still need to have a human in the know to actually look at the case.”
Relying entirely on the results of the model could indeed backfire. Hauge reported that four years ago, the Dutch tax and customs administration wrongly accused 26,000 families of child benefit fraud. The Dutch government resigned over the scandal.
The consequences of poor AI for music selection on Spotify are much less than if it were used in the financial sector or after a medical examination.