Artificial Intelligence, So Mom Can Understand

Dear Mom,

Artificial Intelligence
Artificial Intelligence, So Mom Can Understand


Even though my work often mystified you, you and Dad have been my biggest fans throughout my over 25-year data management career. Not surprising, as you were both dental professionals, not data professionals!

But as you started hearing about data in the media more and more, your interest piqued to better understand it. I’ve thoroughly enjoyed writing my letters to you over the past few years covering a range of topics including Data management, Metadata, and Data Governance, as well as several industry trends like Big Data, the Internet of Things, Cloud computing and my last letter discussing Digital Transformation.

When discussing technology today, there’s likely nothing more exciting – or more complicated – than artificial intelligence, or AI.

I know you’ve heard the term AI before. Science fiction – and fact – has been talking about machines learning and thinking for themselves for almost 100 years. In sci-fi, it’s often in the form of robots turning against humanity once they’re smart enough to do so! It’s true that modern visionaries like Stephen Hawking and Elon Musk provide warnings about the risks of AI, but I would argue that many new technologies could pose a risk if not used and governed appropriately. But I’d like my letter to you to focus on what AI means to us today, and more specifically, what it means for data management professionals like myself.

At its simplest, AI efforts teach machines to learn so they can advise us.

You may often hear the terms machine learning (ML) and artificial intelligence used interchangeably, and there’s a reason for that. Like humans, machines cannot be intelligent until they learn. The decades of research and innovation in AI has focused on teaching computers how to learn for themselves from their experiences – without having a programmer manually telling the computer everything it needs to know for every scenario or context. Let’s use any shopping experience you’ve ever had as an example.

In the old world (whether you were shopping in person or over the phone), let’s say you expressed interest in one item –a coffee machine. It was up to the salesperson or call center agent you were speaking with to know enough to pitch any other relevant products that you may also have been interested in (e.g. coffee filters, mugs – upgrade to the one with the built-in grinder). The goal of these recommendations was to increase your order size, make it a great shopping experience for you, and ultimately make more money for their company. The only way the salespeople knew the best items to suggest was if their company trained them on specific items to cross-sell (if the training was accurate and remained relevant over time), or from their own personal experience. It was hit or miss for them.

Now fast-forward to today. When you go to and select a coffee machine, Amazon suggests products that are frequently bought together and packages them for you. Amazon also suggests what previous buyers of that coffee machine have also purchased. And guess what – you go back to Amazon a week later (or an hour later) – and you may see a completely different set of suggestions. Amazon uses continuous machine learning to teach its recommendation engine what products to put in front of customers across its entire marketplace. Retailers from around the globe are also investing and innovating with similar recommendation engines to enhance their online and in-person customer engagement, and ultimately, improve their sales.

But of course, AI and ML are not limited to retail experiences. They’re popping up everywhere from in-home assistants (if you don’t believe me, ask Siri, Alexa, or Google Home), to autonomous vehicles, medical treatment, utility monitoring and many other industries and consumer experiences.

Great, AI sounds cool. What does it have to do with data?

Data is the mechanism by which anyone – or anything – learns. The more data available, from the widest variety of sources, as fast as you can consume it, the more likely you will come up with an accurate, useful, timely and trusted decision.

Not all data is equally useful though – without the right context, it can be misunderstood or misinterpreted. For example, it’s a fact that an address in Pearl River, NY has something to do with you. But is this your current home address where a business should ship the product you ordered? Is it your billing address? In reality, you and I both know that it’s your work address from over 20 years ago. Without the necessary metadata describing the type and age of this address, it provides minimal value for any business trying to understand, and effectively engage with you today.

Therefore, a constant stream and wide variety of metadata and data is required to feed any machine learning engine to provide truly personalized and relevant AI. (You got it – Big Data and IoT can play a huge role here!)

The opportunity for AI to transform the business world, along with my data management profession, is equally as compelling as our experiences as a consumer. By 2020 Gartner predicts algorithms, like those used in ML and AI, will positively alter the behavior of over 1 billion global workers.

Data management professionals, whether they are in IT roles building data-driven  solutions, or business roles responsible for governing, maintaining or analyzing data, spend a significant amount of their time aggregating, reconciling, cleansing, and otherwise preparing data just to ensure it can be available and trusted for use. I’m incredibly excited that my company, Informatica, recently introduced our CLAIRE AI engine. CLAIRE uses AI and ML techniques, powered by enterprise-wide data and metadata, to significantly boost productivity of all managers and users of data across our customer’s organizations. CLAIRE will allow our customers to deploy data solutions more quickly, with higher confidence, and will help to predict potential issues or automatically resolve them. And best of all, CLAIRE will advise next-best actions and recommendations for all types of data users to help them deliver great results.

It’s an exciting time to be in the world of data, and I’m optimistic that we still have time to figure out how to keep our AI-fueled machines on a leash, while we unleash the power of data for the good of all!

Love, Rob


P.S Tell Dad I love that he’s an early adopter of new technology, and that he proudly talks to Siri on his iPhone, Alexa on his Amazon Echo and Google on his Google Home. But no, he can’t ask Siri to tell Google to order anything from Alexa.