That’s an intelligent question that nobody should be afraid to ask.
Here’s a reasonable working definition to start off with: ‘AI is the ability of a computer or machine to do what previously only humans could do.’ After all, machines have been doing stuff we can do - walking, running, hauling, writing, for centuries. But better. Think of the wheel. Or steam engines. Printing presses. Automobiles. All working harder and more powerfully to replicate human activities, but faster and more consistently. Magnifying what we can achieve, by going way beyond our natural physical strength or mental stamina.
Lots of people still have a sense of ‘it’s coming’ around AI. But that’s not accurate. It’s already here, and it has been for a while. Every time you interact with a chatbot, or you get a prompt when you’re doing an online search, you are in fact engaging with artificial intelligence. You may have used DALL-E to create images based on your verbal instructions. Again, you’re directly engaging with AI-driven technology. You might be amazed by it. In the case of some chatbots, you might be amused or downright exasperated. But you haven’t been replaced by it, and you won’t be!
There is no ‘single thing’ that makes up AI. That’s important to understand as part of demystification. We’re not dealing here with a ‘force’ or a ‘power’. Instead, what we have is an array of technologies that have evolved to a point where they can help us in more ways, and with more sophistication, than at any time in our history. BUT ... they have building blocks that we can describe and understand; in the same way that we can describe and understand what happens under the hood of our automobile.
Machine Learning - ML – is the foundational component of much of AI. In fact, other AI components, from chatbots to image creators, are built on specific types of machine learning. ML essentially means training a machine with data, so it can be used to do a whole variety of different things and power many different applications. It doesn’t give a computer, or its software, ‘a mind of its own’. What it does do, combined with other AI components, such as Expert Systems, is create an artificial capability to recognize patterns and draw insights and conclusions from them. Processing very high data volumes at very high speeds. And getting even ‘smarter’ at spotting issues, the more processing it does.
Augmentation and enhancement are the key themes here. AI, or specific underlying components, can be configured, for example, to create an AP, P-Card or T&E system that can apply thousands of rules (if necessary) against a piece of financial data. Or against thousands, or thousands of thousands, of pieces of data. The rules-based decisions embedded in Expert Systems (one of those components of AI deployed in this example) are like having instant access to the expertise of hundreds of forensic auditors and finance professionals.
The intelligence applied may be artificial, based on computational rules. But it is derived from human expertise and applied in the service of human need for insight, accuracy, and risk and loss reduction. Humans could theoretically do the same job. But there would need to be hundreds or thousands of them in place. AI provides augmentation and enhancement of human capability, in the same way that an automobile enables us to travel faster than we could ever run.
Join Oversight’s Nathanael L’Heureux, Chief Solutions Ambassador and elevate your career in finance by watching his three-part video series, ‘Becoming an AI Champion’. Nathanael tells you everything you need to know to demystify AI. And it’s completely FREE!