This article will provide an overview of artificial intelligence, what it is, and examples of the use of artificial intelligence in finance.

What Is Artificial Intelligence?

There is no universally accepted or standard definition of artificial intelligence, but a commonly accepted definition describes it as “machines that respond to stimulation consistent with traditional responses from humans, given the human capacity for contemplation, judgment, and intention.”

How AI Works

Artificial intelligence comes in different forms, but AI is a general ability to use real-time data to make a decision. The machine or program can receive that data through sensors, remote input, or digitally. The AI then must analyze the data before making a decision, which is the characteristic that differentiates it from a pre-programmed machine. In finance, artificial intelligence can be used in the underwriting process to help a lender make better decisions regarding loan applications. Rather than rely on predictive analytics prescribed by statisticians, a computer algorithm can read data on prior loans and determine for itself the best predictive model to assess the creditworthiness of applicants. Robo-advisors are another popular use of artificial intelligence in finance. Robo-advisors use client information about financial goals, risk tolerance, and investment horizon to determine an investment asset allocation. The robo-advisor then rebalances the portfolio as needed, placing trades and even handling tasks like tax-loss harvesting.

Types of Artificial Intelligence

In general, there are four broad categories of artificial intelligence: reactive, limited memory, theory of mind, and self-aware. Think of these types as a progressive spectrum; each type builds on the complexity of the type before it.

Reactive

This is the most basic type of AI. Purely reactive artificial intelligence can act based on an assessment of the current situation but is unable to build a repository of memories to draw from in the future.

Limited Memory

Building on the reactive category, limited-memory AI can “remember” past experiences as pre-programmed representations of its environment. Limited-memory AI will then incorporate these memories into future decisions.

Theory of Mind

This type of AI is even more advanced than limited memory. Taking its name from the psychological term, theory-of-mind AI can attribute mental states such as beliefs, intentions, desire, emotions, and knowledge to others. If that sounds futuristic, that’s because it is. This type of artificial intelligence has yet to be developed.

Self-Aware

Going beyond theory-of-mind AI, self-aware AI has the ability to form representations about itself—thus having consciousness.

Artificial Intelligence vs. Machine Learning

Because of the lack of a standardized definition and the fact that there are so many related terms, it can be difficult to distinguish between artificial intelligence and machine learning. Artificial intelligence is a broad term, and it’s loosely defined. Machine learning is a particular application of artificial intelligence in which machines learn from data and change over time to make better decisions about that data. The main use of machine learning is to process large amounts of information in a short amount of time. An example of machine learning is the way social media platforms learn what type of content—posts and ads—that you will like more based on how you have interacted with content on the platform.