RiskTech Forum

Wolters Kluwer: Artificial Intelligence Everyday

Posted: 1 September 2014  |  Author: Pim Nauts   |  Source: Wolters Kluwer

Perhaps the most important reason I love summer is it always seems there is so much more time. Time for vacations and to read – to follow every little curiosity and association.

Back in university, they taught me a thing or two about artificial intelligence, the science of modelling human behavior in machines. So I picked up some books from “back in the day” (unfortunately, they didn’t teach us how to escape time). It brought me back to class where we learned about algorithms (really just a string of operations to transform a datapoint), machine learning (can we go beyond instructing a machine and instruct it how to instruct itself?) and human language processing (think voice input, and natural language search). Artificial Intelligence (AI) really is just a way to have machines do the things humans can, only several magnitudes faster and scalable.

Being away from the world of unlimited curiosity for some time now, it suddenly struck me how much of the technologies and applications we discussed in class are now everywhere around us (and how impressive they’ve become). It’s a pity that a field which contributes so much to everyday comfort and safety is so hidden under the surface. So, what everyday AI is hiding in plain sight?

Virtually every online service is driven by AI
In ecommerce, you cannot not come across AI – every recommendation for items you get is determined by algorithms that look at the topics and interest graph you show as compared to other people (“People also liked…”). The ads you click in search results or on websites are bought via automated auctioning, with just a few human-defined variables (like price). Airlines make sure they charge the optimal price for the seat you book online using an incredible array of input signals.

Even my lawn could use some AI
But perhaps, in ecommerce, you expect AI or some sophistication to be there. But it’s everywhere. I often speak to people who think of AI as something “scary.” Well, in that case: good luck escaping it.

When I look at my pay check, what I contribute to our pension fund is (hopefully) multiplied by automated trading algorithms (don’t let me down, economy). Driving home, the dynamic speeds on the digital signs are calculated to optimise traffic and flow. When you’re hospitalized, your doctor will use classifiers to help diagnose your condition. I’m saving up for a lawn mower (which is really dumb, but convenient, artificial intelligence), and I hope to find my way into a system that knows when to water the lawn based on weather data.

AI at Wolters Kluwer
In our own operations, AI is everywhere. I receive an app notification if our online revenues don’t meet the (computer-defined) expectations. Every search our clients run on our legal databases is put through an incredible number of operations to define the relevance of every article we have to their specific query. We have tooling to make drafting contracts faster and more precise (with decision trees – AI!). Analyzing legal risks with legal management software is driven by AI, as is forecasting cashflow with accounting software. Every book our clients order online is picked up by some pretty smart robots back in the warehouse, and the computers that define the routes our delivery partners take are some heavily smart and large-scale machinery.

Back to earth
So, artificial intelligence really is everywhere. But it’s nowhere without human intelligence defining it. Just ask a 4-year old and a computer to tell a tomato and a banana apart and see who performs best.