Can we educate every child "to the max"?
Conversations about AI's potential with Google's CEO, a cautionary tale from LAUSD, and more.
It’s a holiday week, so I will keep this short. Mostly, I wanted to encourage you to watch this World Science Festival interview with Eric Schmidt, decade-long CEO of Google and philanthropist at Schmidt Futures, and host Brian Greene.
The interview is LONG, but fascinating. Schmidt has a lot of things to say about the development and use of AI that are important for anyone running an organization or setting policy to hear. Around 1:05 in the video, the conversation turns to education, and things get really interesting and provocative.
Brian Greene frames an education question for Eric Schmidt, noting that we’ve been “teaching the same way for decades” but that “now we have the capacity to really personalize all of this, to create an educational format that is able to modify itself based on learning habits.” He asks, “Is this the future for how we’re going to teach the young?”
Schmidt responds by first bemoaning the state of data in education and asking why schools of education aren’t producing data on how students learn (dear friends in ed schools and education research: Please overcome your sense of insult when Schmidt questions what schools of education are doing if they are not collecting data on how students learn, and listen on). Schmidt continues by presenting a pretty compelling case for why we should be amassing this data:
The goal should be the following: An AI tutor for anyone in the world, at any level of education. In their language. For free. On their phone. And the AI tutor would adapt to that person’s learning abilities, learning style, attention span, whatever. In whatever is the optimal way. And furthermore, because it has an outcome function, which is learning, it can then go back and solve and modify its own algorithms to get more performance….It looks to me that that product is relatively easy to build if you have the data…Think about a world where every single human is educated to the maximum that they can be. That’s got to be good. It’s got to be an improvement to have more education globally, everywhere.
Greene then talks about the ways that VR systems can allow students to experience knowledge firsthand, like what life is like near the speed of light: “Sure, you can learn things, watch videos, but if you can immerse yourself, at least for some students, you can learn it at a real visceral level.”
Greene and Schmidt then riff on the idea of combining a personalized AI tutor with the VR tools that might allow a student to experience firsthand a concept the AI tutor is trying to teach. Greene: “That to me feels, like, revolutionary.”
Schmidt describes how VR integrations may also be the future of medicine:
Why is it that I don’t have a screen when I walk in with a picture being generated dynamically about what the doctor is talking about? My spine, my hip, etc. Now the doctor is highly trained and has reference pictures. But why can’t they generate my picture?
So when you start thinking about the power of imagery for learning, somebody who gets this is going to build a learning system that’s like a different paradigm that really works…And it’ll take a decade. Of course, the educational system is extremely slow moving, heavily unionized, very resistant to change. It’ll take a while.
But in our lifetimes we should set a goal of having every single person in the world having access to an AI doctor and an AI teacher. We have teachers, but the teachers are overloaded. Health care professionals are overloaded. This would bring everyone up. Everyone in the world has health problems and educational problems. Why can’t we solve them all, now?
You should watch the whole thing, but at least watch the education discussion. The conversation will no doubt inspire a lot of skepticism and possibly defensiveness. I take issue with this pure vision, myself. I worry about kids who don’t have the drive or the parental support to take advantage of these tools. I also worry about how much these tools will cost, who will be able to afford them, and how we can “keep humans in the loop,” as Ethan Mollick would argue we must.
But the fact is, these tools are coming. Fast. Soon. There is no stopping them. For that reason, we in education must listen to these provocations, engage in serious debate, and act on the implications. If we ignore AI tools, inequities could grow, and, even worse, we might miss the opportunity to achieve the positive vision Schmidt described—the possibility of “every person being educated to the maximum that they can be.”
New Research
A few recent research reports help inform how far we are from Schmidt’s vision—or, in one case, how we may be getting closer.
An early study shows that, like human teachers, AI language models can “learn by teaching.”
Another research study finds that we’re pretty far from machines showing empathy.
And another study shows that AI bias should continue to concern us.
District AI Happenings
Here’s a cautionary tale from LAUSD about building big new AI projects in a dynamic ed tech market. The lede:
The future of Los Angeles Unified School District’s heavily hyped $6 million artificial intelligence chatbot was uncertain after the tech firm the district hired to build the tool shed most of its employees and its founder left her job.
The New York Times has a lot to say on the matter as well.
New Uses
How do you find this Sora-generated ad for Toys“R”Us? Creepy? Brilliant? (Editor’s note: yeah, it’s creepy.)
Final Thought
“This thing (Generative AI) is going to happen. Get ready. Try to figure out how to shape it into your institutions so you benefit from it. The next generation of winners will adopt the new technology.”
- Eric Schmidt
Happy Fourth of July! 🎆
-R