Moving at the Speed of Creativity by Wesley Fryer

Stephen Wolfram on Computational Thinking

These are my reflections and takeaways from a captivating 40-minute fireside chat with Stephen Wolfram, shared on August 4, 2023 during the AI x Education Conference streamed via Zoom. Stephen is the “CEO of the software company Wolfram Research where he works as chief designer of Mathematica and the Wolfram Alpha answer engine,” which are two web-based computational platforms I want to understand better as an educator.1

As a middle school STEM teacher, I am particularly interested in Stephen’s take on “computational thinking.” This is my paraphrase of how I understood him to define it: Inherently algorithmic, yet more than just algorithms—it’s a unique perspective through which we can approach any topic or subject. By employing a computational lens, we can innovatively pose questions, paving the way for the generation of data, graphs, and other valuable analytic information which relate to a question or topic. This information is ripe for computational manipulation and visualization. The QUESTIONS we formulate at the outset of an investigation are key, however.

This aligns with some of the pedagogy of inquiry-based / project-based learning which I’ve heard through the years. Helping our students imagine and formulate their own questions about a given topic or content area is even more important than having them successfully “learn” specific, discrete elements of content.

Stephen asserted that the most proficient “prompt engineers” working with large language models (LLMs) like ChatGPT today, demonstrate excellent “expository writing skills.” He sees the evolution of LLMs as a pivotal moment, reigniting dreams from the 1950s—dreams of a “thinking machine.” The patience demonstrated by LLMs, particularly when faced with misunderstandings, he highlights as valuable. While they aren’t without their flaws, like the occasional hallucination, he gave the impression he thinks those issues are ones we can work with and around in many contexts.

Stephen emphasized that in our age of rapidly advancing automation, the value of “a generalist” has skyrocketed. It’s those individuals who possess broad knowledge and the ability to think critically across domains that are becoming indispensable. Narrow skills, while useful, are arguably not as important for students in today’s classrooms as the ability to demonstrate flexibility, adaptability, and the capability to learn how to use new tools on the fly.

These insights inspired me to wonder how I might encourage my students this year, especially those in my coding classes, to delve deeper into data representation and visualization. I want to provide opportunities for my students to develop skills within the “analytics” arena of our “IDE@S department.” (IDE@S = Innovation, Design, Entrepreneurship, Analytics and Sustainability.)

Interestingly, Stephen championed the idea of computational thinking as a liberal art, intricately intertwined with philosophy. It’s about introspection, the act of “thinking about thinking,” and cultivating the skill to ask probing questions and analyze responses critically.

On a side note, I’ve added to my reading list Stephen’s book, “An Elementary Introduction to The Wolfram Language.” It’s available online for free and provides “a non-mathematical approach to computational thinking.”

Stephen also shared a practical use-case for LLMs. He currently employs them to distill a weekly, linked list of 25 news articles about research on LLMs, handpicked by humans, into concise summaries. He finds this to be more efficient than sifting through abstracts penned by the original authors.

One anecdote that particularly stood out was about LLMs crafting “hallucinated” functions for Wolfram Alpha. Intriguingly, their developers found one of these ideas so groundbreaking that they’re now in the process of implementing it—a testament to the unexpected creativity these models can exhibit.

AI Attribution: I used Apple’s iOS speech-to-text technology to dictate an initial draft of this blog post, then used ChatGPT 4 to refine and improve that narrative article. I further edited that result to create this final blog post.

1 – Stephen Wolfram. (2023). In Wikipedia.

Stephen Wolfram on Computational Thinking” (CC BY 2.0) by Wesley Fryer

If you enjoyed this post and found it useful, subscribe to Wes’ free newsletter. Check out Wes’ video tutorial library, “Playing with Media.” Information about more ways to learn with Dr. Wesley Fryer are available on

On this day..



, , , ,