Knowledge Should Be Shared

Like the Founding Fathers of the United States, I find certain truths to be self-evident. The sun rises from the East. Mexican food tastes better in Mexico. And, knowledge should be shared.

While my take on Mexican food may not be held universally, I always assumed the other two were…until this week. I am reading “Genius Makers” by Cade Metz, a non-fiction book on the evolution of Artificial Intelligence (AI) through the stories of the academics that made it happen. In Chapter 2, the author references two occasions where scientists declined to explain a discovery in layman’s terms. One of the examples is Nobel Prize winner, Richard Feynman, who said “If I could explain it to the average person, it wouldn’t have been worth the Nobel Prize.” I’m sorry, WHAT?!

Several years ago I participated in a communications workshop with an organization called COMPASS. COMPASS was founded with the mission “to help ocean scientists effectively share their knowledge in the public discourse and decision-making.” They believe, like I do, that collective understanding of technical subject matter is critical to amplifying the impact of scientific work. To that end, they have shared their expertise in strategic communication with 10,000+ scientific experts in workshops like the one I attended.

Now, I don’t know the underlying motivations or context behind Feynman’s quip — or a similar one from Geoffrey Hinton, the “Godfather of AI,” who invoked Feynman when he declined to explain the Boltzmann machine in a way that people who are not adept at math and science can understand. But their attitude sets a dangerous precedent. COMPASS put it best: science doesn’t exist in a vacuum. It needs society. Scientists have to be able to explain their work and its impact to build trust in science and enable the application of their findings. What good are quantum electrodynamics and the Boltzmann machine if they can only be understood by a select few?

Coincidentally, I have found hours of interviews in which Hinton breaks down his work, including one with “The Diary Of A CEO” that I found particularly easy to follow. So, it’s not like he’s hiding his research from the masses. But his and Feynman’s open refusal to simplify science struck a chord because I strongly believe it can and must be done. It’s a hard but critical step in the scientific process.

Feynman’s phrasing in particular made me think of a challenge my colleagues and I face daily. Technical subject matter is complex and nuanced. Throughout my career, I have seen experts (including myself) receive feedback that we’re diving too deep into the weeds, or sacrificing readability for accuracy. Our biggest challenge is knowing that nuances exist, and letting them go temporarily to allow the broader message land. We don’t always get it right, but at least we’re trying.

Whether Feynman and Hinton’s comments stemmed from perfectionism, intellectual laziness, hubris or something else, today I am disappointed in some of our most celebrated scientists. Their flippant attitude to making their discoveries more accessible to the masses gives the impression that the average person is not smart enough to understand complicated topics so why try? It also gives young scientists the impression they don’t need to put in effort to expand understanding of their work beyond a small circle of their peers. But, knowledge should be shared.

What do you think?

(P.S. Ironically, it was ChatGPT who helped me understand what the Boltzmann machine is and what it does high level. It combines the architecture of a neural network with the statistical mechanics of a mathematical model, laying the foundation for unsupervised learning and ultimately, modern deep learning used in AI systems today.)