After attending this year’s HR Tech World in Amsterdam, journalist Phil Wainwright made an interesting observation about a trend amongst product companies. He explained that they’re layering in a superficial layer of artificial intelligence (AI) — e.g., an Alexa skill — into their products just to be able to claim that their product uses AI. He calls this trend “Machine Washing.”
I’ve spent my entire career as a data scientist watching some “real” data scientists look down their noses at other “fake” data scientists. I’ve even seen Foxworthy-esque articles informing people that they “might be a fake data scientist if…” Ultimately, none of this is particularly helpful. In fact, I’ve seen it generate a fair bit of consternation amongst business leaders and executives who are actively trying to figure out what all of this AI stuff is about in the first place.
There are two forces in the marketplace today. One voice says, “Anyone can do this with the help of advanced tooling,” and the other voice says, “Don’t listen to those guys. You should hire expensive consultants and PhD’s instead.” Laying aside the arguments about which one of these voices is right (hint: it’s the first one), the truth is that it doesn’t matter whether your AI is real or fake. What matters is making progress.
You’ve got to start somewhere
I work for a software company that has built an automated machine learning platform. I’ve spent the last 2+ years working with business analysts and MBA’s to build predictive models. One of the things that we realized early on is that there’s a massive amount of confusion about what AI even is. AI makes some people think about robotic process automation. Others think about Siri-esque services. Most people don’t really know what to make of it.
We started offering a course called Data Science, Machine Learning, and AI for Executives a while back, and it’s been very successful. Basically, we’re trying to teach business leaders three things: First, what do all these buzz words actually mean. Second, how do you spot opportunities to use AI in your business. Third, how do I use AI to build up a competitive advantage.
One of the key things that we teach people is that it’s not about finding “the right” use case. It’s about identifying tons of potential opportunities and then executing on as many of them as possible. Whether or not the opportunities are “real” AI or “fake” AI isn’t relevant — only whether or not they impact revenue and organizational success.
Small changes are sometimes the most important changes
As a frequent traveler, I use Uber a lot. It seems that there are always yellow cabs around me, but I just stand there and wait for my Uber. Why do I do this? Because the Uber app handles the payment transaction seamlessly. I prefer google docs to Microsoft office. Why? Because sharing and versioning is 100X better with google docs. I use the Mac mail client instead of the gmail web client for my mail. Why? Because the user experience is better. As Steve Jobs put it, “You have to start with the customer experience and work backwards to the technology.”
It’s strange that the superficial features of a product are often the ones that make the difference between adoption and failure, but it makes complete sense. Part of me was surprised to hear this “Machine Washing” criticism. The other part of me is resigned to it. As a user of many different types of software — including the software that my company makes — the user experience is easily the most important, most visible aspect of any piece of software. Reducing user friction should be the first thing in the minds of every product company in the world. The best software in the world will fail if it’s designed poorly, and the simplest, most rudimentary software in the world can be life-changing if it’s designed right.
The same holds for AI adoption. It’s not about spending millions of dollars to revolutionize the company. It’s about making many small changes that compound over time. You could:
- Change the way your company handles sales prospects by ranking them with a machine learning model.
- Enhance the way people interact with your product by adding voice assistance.
- Improve the way your organization maintains its equipment using predictive maintenance.
- Optimize the way you organization sets sales targets by predicting pipeline for the coming year
- Reduce customer attrition by identifying at-risk customers with AI
An organization doesn’t become AI-driven by making a big investment, hiring an army of people, or writing a gigantic check. An organization becomes AI-driven by looking at every part of its business— no matter how small — and looking for ways that advanced technologies can improve operations and profitability. Organizations that invest in these small changes will one day look around and realize that their business operates more efficiently, suffers fewer losses, and creates a higher ROI than any of their competitors.
Since I included a Steve Jobs quote earlier, I suppose it’s only fair to close with an Elon Musk quote: “If your competitor is rushing to build AI and you don't, it will crush you.”
He’s not wrong.