How generative AI can help business leaders to think better, according to Coursera CEO Jeff Maggioncalda

In his leadership and CEO courses, Jeff Maggioncalda, CEO of online learning company Coursera tells business leaders not to expect technologists to know where the most valuable customer problems are – and that’s never been more important than now with the generative AI hype cycle at full pelt.  Maggioncalda says:

You have to start with business value and then make sure the technologists are on the team. People haven’t quite realized it’s not just  about the builders. It’s about the users. You want to get your builders and users in the same room to figure out how to solve customer problems. It’s not that complicated, but I think it’s a bit of an organizational habit and blind spot that leaders often just go, ‘Hey, this is a new AI. Go, talk to the people that do technology’.”

As an online learning platform, Coursera offers about 800 AI courses, of which over 350 of them are for generative AI. Maggioncalda says that the company has four or five enrolments per minute onto generative AI courses globally, and up take in some cases has increased by 1000% in the last couple of years. He says:

The different thing about generative AI is that it is not just for the machine learning statistical engineer who builds the model. Gen AI courses are for everybody else who will use it. These are skills on how to use it for marketing, software development, for data analysis and they’re not just content, they are credentials. Increasingly, people are going to have to not only learn jobs better, but they’re going to have to prove that they have the skills.”

Maggioncalda firmly believes business leaders have to get really close to the builders, and that means learning about the technology, and understanding what it is capable of, and what it isn’t, notoing:

What I often see is that businesses will look at the technology and say, ‘Hey, figure out what we should do with this technology’. And then what the technologists do is they tinker with the technology. They’ll fine tune it, they’ll play with the different algorithms. They’ll move the temperature, they’ll explore different models. What I say in my course, and what many users are saying too, is, first focus on the most valuable problems to solve for your customers and your business. Then ask what these AI capabilities can achieve.

He adds:

The capabilities are not infinite. AI can’t do everything right now, but it can do certain things quite well. Understand enough about the domain of feasibility and capability, and enough about the domain of value and problems to be solved, and then find the intersection.”

AI dog food

Coursera itself has ‘eaten its own dog food’, using generative AI to tackle a list of things it wanted to provide for customers. For example, language translation. It wanted its courses to be available to every single person globally, no matter what language they spoke. Maggioncalda explains:

Not just written, but written, audio and video. And right now, courses are available in 21 languages because we used AI. But that was a problem that was sitting out there for a long time, and we saw the highway and said, ‘We can solve that problem now’.

One example is the firm’s Coach offering, which provides interactive online help. Maggioncalda says:

We wanted to actually scale not just the videos, but the student support in helping explain things, and helping them practice? And that was a problem that was sitting out there for a long time we dreamed of solving.

The company used generative AI to develop Course Builder for instructors. Maggioncalda explains:

It takes a long time to build these courses, and we thought generative AI might allow an instructor to build courses very rapidly. By mixing and matching the best of multiple courses into a single course, they can meet their specific learning objectives and audience needs.

Demand for job skills is dictating the content Coursera is putting into its generative AI courses. Courses for jobs in data science, technology and business leadership, and management are all in high demand, according to Maggioncalda:

For those jobs, we look at what skills are needed, not just the foundational skills, but some of the immersion skills. And then once we know the jobs we want to focus on, and the skills that are needed, then we look at who can teach this course the best, and how.

Partnering

To develop its courses, the company works with around 325 universities and industry partners globally. Maggioncalda says that in the early stages, onlines platforms provided passive learning, then active learning, but generative AI has led to an interactive approach:

You’d watch a video on YouTube, you just sit there and watch it like you’re watching Netflix. You don’t do anything. And then Coursera and others brought in a new era of active learning online. You watch the video and then you do something, you build a project, you answer a question, you take a quiz, you reflect or discuss it with your peers. What generative AI is going to do is move from active learning to interactive learning. Coach is like a tutor who is there the entire time you’re learning, and is summarizing things for you. It’s explaining things for you. It’s showing you how what you’re learning is relevant to your job.

Maggioncalda believes that in the future generative AI tools will offer  CEOs, business leaders and employees alike a safe space to think and try out different ideas, so it is important that they understand what it can offer. He concludes:

For anyone who thinks, and anyone who communicates thoughts, which is everybody, then generative AI will really be a tool that can help you think better.

My take

With all the hype and noise currently surrounding generative AI, it makes sense for decision makers to take the time to learn about and really understand what the technology can offer. Businesses have never been that great at bringing together users and technologists to capitalize on technology developments, but if Maggioncalda is right, this could be about to change.