With AI, the Biggest Risk Isn’t Moving Too Fast. It’s Standing Still

Everyone today wants to be "risk aware."

I’ve heard that two-word phrase a lot lately in a variety of settings. Be bold, rather than timid, we tell ourselves, our colleagues and others. Have courage, rather than fear. Have a watchful eye on potential hazards and negative impacts, of course, but step forward, rather than back.

Let’s call it "smartly bold."

We all want to be that way, don’t we? But let’s be honest. We live in a risky world, and it is scary out there. Many of us are more risk-averse than we care to admit, a little more chicken than we might want to be (no offense meant to the chickens). After all, who wants to get hurt?

Among the many places where I see the struggle to find a balance between risk aversion and awareness is in the realm of artificial intelligence (AI). The advent, explosion and opportunities of AI seem to be changing everything everywhere all the time these days. The potential for good is great; the potential for harm is its equal.

That certainly is the case in the field of education. AI has the potential to transform education in a positive way. However, it is also seen as a doorway to increased cheating as students lean on AI to write papers, take exams and trade its use for their own learning. Interestingly, early research doesn’t really support that fear. Using an AI detection tool, a study by EdWeek.org and Stanford University found that only three of every 100 student writing assignments surveyed were generated mostly by AI.

"We know from our research that when students do cheat, it’s typically for reasons that have very little to do with their access to technology," said Denise Pope, a senior lecturer at Stanford Graduate School of Education.

We need a risk-aware stance to lean into AI to accelerate learning. That’s why I’m excited by the work my UNICEF education colleagues are doing to embrace AI. Here is the most recent example: 

In a recently announced partnership, OpenAI and UNICEF are working together to leverage Generative Artificial Intelligence (GenAI) to convert learning materials quickly and efficiently to increase accessibility for children with a range of disabilities. The collaboration focuses on a two-year pilot program in Uruguay to create an AI tool to do this important work. Other partners include government agencies, organizations of persons with disabilities and curriculum developers.

The UNICEF Venture Fund is also wading into AI-driven educational tools. The Fund makes small investments in early- and growth-stage technologies for children developed by UNICEF country offices or companies in UNICEF program countries. One such investment is with Afrilearn in Nigeria. It has an AI-empowered digital app that helps learners progress from easy to more challenging content to personalize their learning journeys through their national curriculum. So far, more than 30,000 students are using it.

Another investment is in Bookbot in Indonesia, a virtual reading assistant that helps young learners improve their literacy with speech recognition technology that listens to them and then assists them as they read.

And one more: in Burkina Faso, the Venture Fund has invested in U-Youth, a mobile phone-based platform that uses voice-to-text translation to bridge language barriers in a country where multiple local languages are spoken, to give youth a voice on issues they care about and access to information.

These are bold ventures embracing the upside of new AI-powered technologies. It wouldn’t take long to list a variety of risks associated with each. But they are risks to mitigate, not reasons to back away.

I like to think I’m risk aware. My wife isn’t so sure. When I come back from a long bike ride and regale her with stories of going too fast down a long hill, she’ll say I am clearly not aware of the risks. A car might brush too close. A deer could jump out onto the road. Maybe there is a rock or a branch on the pavement ahead.

OK. Sure, I say. I am aware of those risks. But—and don’t tell her this—I don’t really have a mitigation plan.