Ai is Much Simpler Than Most People Think
The Devil is in the Details
I have worked in the data science field for quite a few years. Coming up on 15 to be a bit more precise. It all started in a research laboratory and has blossomed into a career with diverse experiences.
From researcher, to developer, to leader, and now entrepreneur I have seen many aspects of data science, experimented with more potential solutions than I can count, and have the digital scars to prove it 😊
What have all of my mistakes taught me? Well, the short of it is that all of my mistakes have forced me to take a very practical view of artificial intelligence. I mean after all, I was charged with developing Ai strategy from scratch at a company that barely had data scientists.
The pressure to be successful amidst such difficult expectations forced me to come up with a super practical definition of Ai that I could communicate to others to get their buy-in and eventually seek their help in operationalizing.
So, what definition of Ai did I land on?
To be clear. I had read the myriad of definitions that were available from experts, futurists, and technologists. Those definitions were great for an academic audience or perhaps a high-level CEO of a multi-billion dollar company who merely had to throw money at the problem and didn’t have to worry about actually translating those definitions into business value.
I came across definitions that looked like this:
“In general terms, AI refers to a broad field of science encompassing not only computer science but also psychology, philosophy, linguistics, and other areas. AI is concerned with getting computers to do tasks that would normally require human intelligence.” — Deloitte (2017)
Right, so based on this definition a company would need to hire a psychologist, a philosopher, a linguist, a computer scientist, and some nebulous “other” domain expert before trying to enable an Ai practice that could begin to create some business value. Geesh, we haven’t even started and we are already millions in the hole just in personnel.
Maybe I am overdramatizing things a bit but the challenge to create successful Ai practices in business is well documented. Many businesses, particularly small and medium-sized businesses, still struggle with successfully applying Ai in ways that bring clear business value.
That’s why an easy definition is all the more important. So back to the definition I landed on that would help to fuel my success. The super practical and easily operational definition I came up with was the simple idea that Ai was the mere putting together of two things; data science and software applications.
That’s it. No hype. No buzz words. Just find ways to combine data science with software applications, focusing on the ways the improve user experiences, and wham! Artificial Intelligence.
This was the super…no…hyper-simple philosophy that I touted when I first started to build an Ai practice for a large enterprise. It was successful because people understood it. People could relate to it. It made Ai seem much more achievable. Thus, people were far more likely to get behind my efforts and want to support what we were doing.
Is combing data science with software applications a simple task, absolutely not. As the title suggests, the devil is in the details. Just as the idea of business is a ridiculously simple idea (e.g. provide value to people and you will succeed in business) the actual implementation of this simple principle can still be challenging.
Nevertheless, data science embedded in software applications provides those applications with the ability to deal with some amount of uncertainty. And dealing with uncertainty is something that we humans excel at. Thus, laying data science into software applications give the impression that the application can now do things that appear human-like.
As you consider your own DS/ML/Ai journey, remember that despite the hype and media reporting on crazy sophisticated data science models training trillions of parameters with billions of data points, for most use cases the real business value is merely to use data science to improve the experience behind software applications.
Be those applications hidden in machines and robots, or behind GUIs accessible to customers or staff. All we are doing in Ai is helping those applications to use data, either historical or user generated or both, to make predictions, classifications, or forecasts that give it the appearance of human intelligence.
How we make these connections depends on the application, the means of delivering the data science, and the specific use case that helps to define what the value-add is exactly.
Like engaging to learn about data science, career growth, life, or poor business decisions? Sign up for my newsletter here and get a link to my free ebook.