Navier Stokes, AI and Francis Bacon
How is a book from 1620 useful for today’s problems?
II. The unassisted hand and the understanding left to itself possess but little power. Effects are produced by the means of instruments and helps, which the understanding requires no less than the hand; and as instruments either promote or regulate the motion of the hand, so those that are applied to the mind prompt or protect the understanding.
(Francis Bacon, Novum Organum, Aphorisms I.II)
Francis Bacon wrote the Novum Organum in 1620, not long before he fell from power in disgrace. His work defining inductive reasoning has long been influential in its broad strokes; less so than in its details. Still it remains a foundation of the scientific method.
The unassisted hand and the understanding left to itself possess but little power.
Bacon is starting his case for inductive reasoning here, but I think this aphorism also describes why there is a need for something like AI — as many others have said, AI has the potential to be disruptive for thought work in the same way that the Industrial Revolution was disruptive to manual labor.
…instruments and helps, which the understanding requires no less than the hand…
However, the Industrial Revolution lies 150 years ahead of of Bacon’s time. His work and his words are creating a foundation for the possibility of that later event, although it was not inevitable: as the attempts to industrialize the world in the late 20th century show, humans are very good at creating the conditions for an industrialized country to rise; we are less good at actually making it happen. But that’s tangential: my point is that groundwork must be laid.
Get Kyrce Swenson’s stories in your inbox
Join Medium for free to get updates from this writer.
Right now we are trying to instantiate the thinking Industrial Revolution without doing the prework.
…those that are applied to the mind prompt or protect the understanding.
Recently there has been a lot of drama around OpenAI’s solution to Navier-Stokes. Leaving aside the juicy ethics and intellectual theft issues — how often do we get a math-related melodrama!? — I find Terence Tao’s reservations most interesting. Tao feels that by going directly to a solution, without scoping out all the steps on the path to that solution, we are not, in fact helping mathematics move forward. Even in elementary school math, you show your work.
Most AI implementations in the post chatGPT world skip showing the work. AI chat does not “prompt or protect the understanding” without a lot of deliberate work to make it do so; it accelerates our misunderstandings and knowledge gaps unless we deliberately force it to do otherwise.
Bacon’s Novum Organum is a method for a better implementation of AI in the world. The next time you open Claude, or Gemini, ask yourself: am I prompting or protecting my understanding? Or am I removing understanding at all?
Novum Organum from Project Gutenberg https://www.gutenberg.org/ebooks/45988
Wikipedia on Francis Bacon https://en.wikipedia.org/wiki/Francis_Bacon
Terence Tao on Navier-Stokes https://terrytao.wordpress.com/2026/09/07/finite-time-blowup-with-smooth-forcing-term-for-the-incompressible-porous-medium-boussinesq-and-incompressible-euler-equations/
