Making Sense of Generative AI
The mention of artificial intelligence can elicit excitement, worry, confusion, and maybe even a little boredom. Now throw in a new term, “generative AI”, and the feelings start all over again. Underlying it all is a basic question. Aside from producing lots of “cool” information really fast, people want to know what this latest technological development can do to make their work and life better.
As a lecturer at MIT’s Sloan School of Management in managerial communication, Melissa Webster helps students, executives, and employees figure out the answer, because, adding to the confusion, there is no one, single answer. There never is with any technology and generative AI doesn’t buck that trend.
When generative AI was released upon the world, there was no manual with it
There still isn’t, and there likely won’t be. But that’s not a bad thing. It just takes some thinking about what you need, followed by experimenting, being flexible with what you discover, and staying clear-eyed through the process in order to answer the question that might give you your answer, “Am I getting what I want?”
Living with Some Unknown Generative AI is not the first technology to be seen as both a must-have for success and an inevitable threat to employment. It joins a list that includes computers, laptops, cellphones, and the internet and comes with a certain hysteria over the devastation it will cause. With the internet, the prediction was it would ruin the legal industry. It didn’t, Webster says, but 30 years ago no one would have imagined that one day people could take out a cell phone and have a car pick them up where no cash changed hands, or that they could see photos of a stranger’s home and make a reservation to stay in that place while on vacation, again with cash never changing hands.
Regardless of which technology it is, Webster says any new platform follows the same cycle. At first, there’s curiosity. (What’s this?) Then excitement, coupled with some over-hype. (This is the coolest thing ever.) And then boredom. (Oh yeah, that again.) It’s the point when the benefit has become well-established. “There’s a leveling off,” she says.
Generative AI is still in the getting-to-know-you phase, but there’s only so much to learn. Whatever it does now will likely be different in two months, and that fluidity lends to the uncertainty and hesitancy to fully commit to adoption or implementation. What helps, she says, before anyone dives in is to figure out what you’re trying to do and what you might need help with.
If you’re a writer, it might be brainstorming ideas or getting feedback after a first draft. Or maybe for a company, it’s amassing photos or audio files. Or maybe you don’t know what you need. Whatever the situation, the best thing to do is to experiment and test what you discover. You’ll find one of two things: it helps or it doesn’t. If it’s the former, you’ve removed some mystery and anxiety. If it’s the latter, before writing the technology off completely, you first have to do some exploring to see if other people are running into similar issues. “The functionality is often hidden, dependent on your skill to bring it out,” she says.
The challenge is that generative AI is just that, generative. It will give you a lot of whatever you’re asking for, and it can be hard to tell the value. People come with blind spots and it’s hard to know where your personal ones are. Webster suggests to follow a thought leader– and there’s always someone, either in your company or online, who’s tinkered and explored more than you have.
A company needs to do the same thing, experiment. She says that McKinsey did a study and found that successful ones see AI as a way to increase revenue and innovation, not as a means of savings. They also keep a human in the loop, setting tasks, asking questions, and maintaining a critical eye on what’s produced. But the biggest thing a company needs to do is set a culture that openly supports exploration. It’s not just in words. It’s by doing things like holding hackathons and sharing prompts of the week. Without that, offices will develop “secret cyborgs”, employees who will be experimenting but not sharing what they find, because they worry about how it would affect their jobs. The only thing being lost in this scenario is useful information. “You want to benefit from it,” she says.
Executives also need to remember that while the discovery phase can be benign, ultimately, there could be disruption in how people work. Technology has always affected, and sometimes eliminated, jobs, e.g., the elevator operator, ice deliverer, and toll taker. Generative AI holds that potential as it’s threatening knowledge-based occupations. What it will do isn’t clear, but as a company starts understanding the benefits and changes that will take place, it needs to do one thing in particular, communicate clearly.
Get to the Headline This is another area of Webster’s research, and whether it’s giving a conference speech or explaining the ramifications of a new technology, the rules don’t change on how to do it. You have something to say? BLUF – Bottom Line Up Front – she advises. Forget about going in chronological order or building drama. Start with the major news, then pepper in supporting information as needed. It’s easy to want to give every detail, but it belies your audience’s reality.
“Sad to break this to you, they may not be as interested in the topic as you are,” she says. “They’re good people. But they have a lot of other things going on. There’s only so much cognitive space they have in the day, and your service is to use as little of that as possible.”
And even if you’re direct, there’s one more thing to remember. You need to allow them to adjust to whatever has been said. However great you think the future looks, it’s not old news to your audience yet. You need to allow them to catch up to what you’re already comfortable with. It’s an easy trip-up, part of what’s called the “curse of knowledge,” Webster says, “When you’ve learned something, you forget what it was like to not know that thing.”