Building for human potential. It's a big deal.

Sometimes you seek an education, and sometimes a tweet comes along and smacks you upside the head with an education.

We are an ed-tech company peopled by a team who cares about an even playing field for young learners, and we develop the technology to provide one. Our adaptive learning engine adjusts to players’ abilities, meeting readers where they’re at to boost comprehension skills.

Our deeply engaging games that activate a true passion for learning are housed in a tablet or on a laptop, for instance, but former Google CEO Eric Schmidt is thinking a few years ahead when he envisions an AI-powered teddy bear twinned with very young kids. The bear learns as it goes, listening to the child’s questions in order to apply the store of human knowledge to a happy, or quiet, or more educated human, whatever it’s optimized for. “The ability to subtly manipulate the developing mind,” says Schmidt , “It’s a very big deal.”

In the present, our algorithms are clever, but the dataset they pull from is vast, and it isn’t failsafe when it comes to exposing kids to world-views or assumptions that reflect a narrower world.

A very public tweet, which now forms part of our origin story, called out some of our content and shocked us into action. In a humbling public response I wrote that “there was understandably a lot of anger, disbelief, frustration, and hate.”

The tweet and 2021’s The Alignment Problem were a one–two punch. Contrary to the prevailing tech wisdom of faking it till you make it, we stopped. We actually stopped building our machine-learning roadmap, and we pulled entire sections in social studies, historical fiction and dialect from our curriculum. We consulted with literacy teachers and an experts in culture and dialect. We found a partner to help us navigate the thorny problem of building in inclusivity. That partner, NYU, developed a scorecard for content, and how we did on that informed a white paper that our new diversity, equity, and inclusion (DEI) team uses to audit our materials–we have committed to updating this paper annually.

Rating our content on a continuum from culturally responsive to culturally destructive was uncomfortable. The results were … Well, what did they used to say on our report cards? “Room for improvement.”

The scorecard prompted questions like How are people of low socioeconomic status represented? Are people with disabilities represented only in stereotypical contexts? Is omitting phrases like “chubby arms” coddling kids or is it a sensible removal of hurtful language? And where’s that fuzzy gray line?

Whatever you think about wokeness, our efforts to instil our content with unhomogenized ideas around language is about broadening minds. We were burned by that tweet, but our mission remains in place. Eric Schmidt sees the pitfalls with the bear, but he wants the bear. “We’re playing with fire,” he concedes.

“We’re playing with enormous impact … and the reason why I want the bear is that the bear will learn how the kid learns, and the bear will then become the greatest teacher the kid has ever had.”

If our technology meets kids at the point they’re forming critical-thinking skills, drawing their learnings from different, broad contexts that support their development into less biased, more thoughtful humans, then we are firmly in the realm of impact.

Success is not a given. While our learning games adjust to each learner, there’s hard work to be done to ensure that that adjustment doesn’t shut out different linguistic contexts, that it doesn’t narrow their world by tailoring it. It’s a tension no one technology company can solve.

Schmidt’s bear is just a metaphor, but he warns that we have to keep the bear from learning the wrong things, and that reflects his concern that our decisions about early education are about to get way more consequential. We have to get ahead of the problems, he believes, because “the future of the world is completely determined by whether we can get the best human potential out of everybody.” I’m in the business of potential for a reason.

?? ?? ?

This is the 11th story in a twelve-part series that takes you through our journey, from an idea to start-up to fundraising, as we apply technology to level the playing field for all learners. There is no poll this week, but if you would like to listen to the podcast of Eric Schmidt sharing his thinking about the future of AI in learning, you can visit it here , or check out this talk between Sebastian Thrun and Dr. Eric Schmidt where they discuss leadership principles, the future of AI, and personalized learning. It's something worth thinking about (all of us).