In your marketing or business career, AI can give you superpowers. But you can use those superpowers for good or for evil.
There are a lot of ways AI can go wrong. We have a responsibility as marketers, technologists, and corporate leaders to make sure we use AI for good.
1. Make sure you have good data.
AI makes predictions based on data.
For marketers, that data could come from email, CRM, advertising campaigns. Everyone, in every business function, has dozens of data sources they use.
If that data is bad, the prediction will be bad. This does damage to your company, consumers, and your career.
Lead scoring is a practical example of this. AI exists to score new leads that come in. It uses all the existing data you have on people who converted. If accurate, AI can effectively score leads at scale.
The problem is the data you give it. If it's bad data, you'll get bad lead scores. When that happens, salespeople are negatively affected. They see bad scores come out and don't trust the machine as a result. After a couple of days, they stop using the system and go back to human intuition.
This ordeal damages the trust your salespeople have in you. It hurts their financial livelihoods. And it costs the company money.
Not to mention, you still don't have good lead scores.
Bad data leads to bad outcomes all around.
2. Make sure you don't have bias in your data.
Bad data is one thing. Biased data is another.
Data comes from humans. Humans often have bias. As a result, your data can contain different types of bias. When it does, AI makes biased predictions. This damages everyone involved.
You see, data can contain the same human bias that causes adverse decisions or inequalities in the real world. AI tries to use that data to make predictions. But when we use AI to make predictions about consumer behavior, we're often looking at personal data. Things like demographics, religion, race, and other individual qualities that signal consumer preferences.
What happens if this data contains bias in favor or against a particular group? What if it makes outdated or even harmful assumptions about different groups? What if it doesn't even consider certain groups at all?
Your AI systems will end up making predictions that contain those biases.
Think about how this can go wrong using the hiring process as an example.
Let's say a company hasn't been very diverse in its hiring practices. All the company's employees are white.
The company wants to use AI to find top performers to hire. So they take all the data on their best employees from the last five years. Everything you can think of, including demographic data.
They train the machine to identify top performers based on the data. The machine begins to flag which resumes the company should review and which it should reject.
The machine begins to flag all non-white candidates for rejection because they lack a typical pattern among top performers: they don't look like them.
The machine isn't self-aware. It's not malicious. It only knows what information the company gave it. All the machine knows is that non-white candidates don't look like top performers.
It's a simple, fictitious example. But don't think this doesn't happen or can't. In 2019, Apple used AI to determine how large a credit line each applicant for its credit card product should get. Because of bias in the data, it ended up discriminating against women.
3. Make sure you use AI responsibly.
It's not just about the data. We all have responsibilities, too.
With AI, we can learn more about consumers than ever before. We can discover their beliefs, interests, fears, and desires. We can use that information to manipulate them. We can make predictions about their behavior. We can trigger those behaviors in ways that are unethical.
We must focus on using these superpowers for good. There are bad actors out there. Organizations use AI to take shortcuts to hack marketing and affect peoples' emotions and behaviors.
Instead, we need to use these superpowers to drive personalization and convenience for consumers.
If we don't, AI can go very wrong.
If we do, we can create incredible things with AI.
To use AI for good, we can't forget there are people on the other end. When we remember that, we end up doing good for others, ourselves, and our brands.
As Chief Content Officer, Mike Kaput uses content marketing, marketing strategy, and marketing technology to grow and scale traffic, leads, and revenue for Marketing AI Institute. Mike is the co-author of Marketing Artificial Intelligence: AI, Marketing and the Future of Business (Matt Holt Books, 2022). See Mike's full bio.