As more of our data is captured in everyday transactions, artificial intelligence is playing an ever greater role in our online experience, whether or not we realise it. In the first part of this short series, we’re clarifying how AI differs from human thinking, and we’re looking at how different industries make use of web-based AI.
What is AI?
The AI that these websites and applications are using to crunch user data doesn’t behave like human intelligence exactly. It aggregates huge amounts of information and seeks relationships and patterns—more information and faster than a human could ever do—but it isn’t capable of drawing upon intuition or lived experience the way people are.
People are capable of improvisational thinking about lots of tasks at once or switching from one task to another without thinking much about it. We are working on one thing when something interrupts us, and we switch over to managing something unrelated. Or we can be in a meeting and something comes up in the course of the discussion that we decide to prioritise over the listed agenda items.
AI, as it currently exists, is necessarily specific to solving the problem it is programmed to solve, very different from people. What it does do, however, is often create an interface that feels like a human interaction. This can be both very helpful and sometimes a little uncanny. AI still seems a bit robotic and stilted when we interact with it. But as it gets more sophisticated, it will learn more natural language patterns, and as we use it even more, we will become more accustomed to its interface.
How AI goes to work
Many industries actively use AI to advance product development, lower overhead costs, or deliver a more precise service. It sometimes appears in their interactions with clients, such as in chat bots or automated marketing on websites. Chat bots use a database of a company’s industry-related information, and learn to recognise the most common ways questions are asked to retrieve answers. Automated marketing uses AI to learn what a customer likes based on their actions and other information about them to sell them new products.
But very often AI works in more behind-the-scenes ways to foster general efficiency. For example, in manufacturing, AI can gather data from different areas of a company—account management and supply chain logistics, for instance—to create a production forecast which is more accurate than what a human can calculate. This saves waste, especially for products with a short shelf life, such as perishable food. By communicating between machines, it can also measure workflow stress, predict machine failure, and optimise maintenance schedules to minimise downtime.
Communication between machines is also being developed in transportation and logistics for self-driving vehicle navigation. These vehicles rely heavily on AI for many functions, including algorithmic decision-making about speed and lane-changing, for example. But communications between vehicles about traffic, road, and weather conditions in real time can also keep things moving even when problems arise.
People power scaled way up
In the financial world, algorithmic decision-making is used to detect tiny fluctuations in market value that AI can exploit to make near-instantaneous trades. Each trade may only be worth pennies, but because the AI is capable of running through so many in a short span of time, they quickly add up. Banks use neural net AI to detect fraud. They train an AI what legitimate transactions look like so that when a transaction occurs that doesn’t match the pattern, it gets flagged.
In a similar way, insurance companies are increasingly using AI to create risk profiles for clients, especially big ones. A logistics company, for instance, with hundreds of drivers to insure. Telematics data, which feeds through from vehicles’ on-board computers and GPS, serves as the database of information from which the AI can determine who deserves a discount. An insurer for a film production or festival may make use of weather forecasting AI for risk mitigation when putting together a policy.
These are just a few examples and a few industries. The fact is, AI is present everywhere the internet is, and understanding how to integrate it into your organisation is going to be a key component of business tech strategy in this decade.
In Part 2, we’ll be looking more specifically at how the power of AI can be put to use in your website and how it affects your customers.
If you’re interested in learning more about AI and how it affects your business sector, contact Technical Director Ben Franklin.