Thinking About Thinking About AI

In my October blog post, I discussed AI’s benefits and pitfalls as they relate to education. Today, I’d like to pivot to another program at the school that is even more important to understand in light of recent advances in AI: Cultures of Thinking.

The Cultures of Thinking website describes the program this way:

We define “Cultures of Thinking” (CoT) as places where a group’s collective as well as individual thinking is valued, visible, and actively promoted as part of the regular, day-to-day experience of all group members. Drawing on previous research by Ron Ritchhart (2002), the CoT project focuses teachers’ attention on the eight cultural forces present in every group learning situation, which act as shapers of the group’s cultural dynamic and consist of language, time, environment, opportunities, routines, modeling, interactions, and expectations.

Cultures of Thinking is a great framework for a number of reasons: it emphasizes thinking over recitation; it includes a great framework for what ‘understanding’ actually means; and it acknowledges the fundamentally social interplay between group and self in learning.

What I most appreciate about the Cultures of Thinking framework is that it emphasizes that some things— traits, dispositions—simply can’t be learned from a textbook or a whiteboard; they must be ‘enculturated.’ They must be “part of the regular, day-to-day experience” of our students to be internalized, after which they should be as natural as breathing. 

One of the habits being formed in the Cultures of Thinking framework is proactive thinking and this is a good example: as much as one learns about thinking in the school curriculum, if you haven’t lived in a school culture where thinking happens all the time, you’re unlikely to walk away energetically thinking for yourself.

To me, this concept extends well beyond thinking: curiosity, multi-modal collaboration, meaning-seeking, willingness to engage with uncertainty, confidence, creativity, complex ethics, perspective taking – these dispositions are more appropriately encultured than taught. One of the reasons why neurodiversity is so strongly supported at THS is because daily perspective taking is good for all students (compare, for example, to a termly trip to see how others think).

Embracing enculturation poses many challenges for schools. It is a harder process. Teachers are forced to look beyond the content of their class, to examine the culture-shaping aspects, such as classroom set up and classroom interactions. It’s also harder for administrators who are required to shape the whole culture – after all, it would be impossible for teachers to enculture students into a classroom culture without working in a likeminded school culture.

Thinking is also slower than recitation. Open-ended questions have far more degrees of freedom than questions with right or wrong answers. This can be frustrating for parents and teachers alike. “Can’t we just tell them about WW2 rather than have them think about it? They’d learn so much more.” It’s hard to consistently deny the temptation of these short-term rewards.

Ultimately, we know the wait is worth it. Our learners walk away with many of the internal dispositions we’d wish for them to have as adults: curiosity, the desire to seek understanding, to develop their understanding, and to be confident in their understanding (and, dare I add, to find meaning in their lives).

In the last blog post, I ended by discussing how AI will help us think about thinking and there is an interesting interplay here. Ethan Mollick of Wharton University, one of the professors leading the way on AI in the classroom has this to say: 

“Right or wrong, writing has been used as a proxy for care & effort. A student who writes 40 pages must be thinking more. A long letter of recommendation means the recommender cares. A government agency producing a 1,000 report means they were thorough. AI breaks that illusion.”

Ethan recognizes that thinking is the ultimate goal. Learning to think is slow, but AI is blazingly fast. The metacognitive skills that THS students are learning are the ones that AI cannot (for the moment) engage in—it’s a harmonious fit. The more sophisticated the thinker (i.e. the more metacognitive), the more they will be able to understand the tools available from AI, seeing clearly where AI supports their thinking and where AI needs their support; now, and in the future. 

Having used AI for more of my work now, I am convinced that we’re looking at a sea change in human potential akin to the Gutenberg Press. Creativity will be unlocked. Deficits will be managed. Thoughts will flow to paper like water. And as a result, collective and individual thinking will be refined.

The Understanding Map