169 years later, Thoreau’s commandment still echoes across the surface of Walden Pond — “Simplify. Simplify!” The logic has stood the test of time. Seriously, could anyone want more complexity in life?
But that doesn’t mean simplifying is easy.
To simplify means to “make something easier to understand,” according to the first online dictionary definition that popped up when I typed the word into the search bar. Understanding takes effort. And time. Both of which are in short supply. So simplifying is a good thing — indeed, it’s the key to solving problems. In this regard, simplify has a similar meaning to compute.
Computation as “Destroying Information”
Using a somewhat quirky turn of phrase, scientists talk about computation as an operation that “destroys information.” I asked ChatGPT what this meant. It replied that in the field of quantum computing, the information contained in qubits is highly sensitive and can be destroyed by external disturbances (which is why quantum computers must be shielded from magnetic fields and operated at close to absolute zero). Then it added that information destruction wasn’t relevant to classical computation. But here it is wrong.(1)
Consider an old-fashioned calculator, like the HP-12C that was popular with engineers and bankers. Type in 4, hit the “enter” button, type in 8 and hit the + button (this peculiar sequence is called Reverse Polish Notation) — and presto, the number 12 appears. However, the inputs (4 and 8) are no longer visible. They’ve been “destroyed,” for the calculation is irreversible, meaning that you cannot recover the inputs from the output (many different pairs of numbers could have produced the same result). With an old-fashioned calculator, you can destroy a lot of information. Imagine adding up 1,000 numbers — at the end, you’re left with nothing but the final summation.
Now to be sure, you could’ve typed the inputs into a spreadsheet column, or for that matter written them down in a notebook of green grid graph paper of the type that accountants used to use, in which case the information would have been preserved. But that’s because you undertook a second operation — that of storing data, which is separate from the computation.
Destroying information is an odd metaphor, but an apt description for how computation solves problems. Consider another example — a linear regression model. You input a large number of observations, and the model produces a simple linear relationship with only two variables — slope and intercept. That equation helps you understand causation. And now you can make predictions.
Adding, subtracting, regressing, calculating, computing — it’s how we figure out the answers to arithmetic puzzles. And solve life’s most pressing problems. The answer doesn’t have to be precise. You can take a guess. You can form a subjective probability estimate using the principles of Bayesian statistics, either intuitively, or by sketching out a probability tree. Or you can use machine learning applications to build a “probability forest” which consists of multiple probability trees. Or you could create a Monte Carlo simulation and run it through millions of scenarios, to better visualize risk.
Simplicity Rules
In our world, data is exploding. Theories are proliferating. Tools are becoming ever more complicated. Yet simplification is always the right path forward. Which is why Einstein cautioned, “if you can’t explain something simply, you don’t understand it well enough.” (Incidentally, I asked ChatGPT if Einstein had actually said this. At first it said yes, but when I challenged it to provide the source, it “apologized for the confusion” and acknowledged there wasn’t one.)
Simplicity is so important because even the most sophisticated computers are mechanical devices, which require time and energy to do their work. For this reason, when reviewing alternative theories, scientists employ “Occam’s Razor,” which means they prefer the theory with the fewest variables. As an 13th century philosopher named Robert Grosseteste commented , “That is better and more valuable which requires fewer [premises]…because it makes us know more quickly.” Occam’s Razor relies on the principal of parsimony. Which the online dictionary describes as the quality of being “thrifty” or “cheap.” As in husbanding your resources.(2)
Think of how long it took our hunter-gatherer ancestors to acquire such deep knowledge of their environment. They moved through the forests, nibbling on different plants, passing down through the generations the knowledge of which species were healthful and which were toxic.
Today we acquire and synthesize data more rapidly. ChatGPT’s training dataset contains around 100 trillion parameters, which would take a high-end processor approximately 355 years to parse. By operating in the cloud, OpenAI, the company that developed ChatGPT, may have accessed the equivalent of 1,200 processors. So perhaps the training was completed in a month.(3)
Computation, whether it takes place in neurons or on silicon chips, is a mechanical process. It takes physical work to reduce large datasets of disorganized information into smaller, more orderly sets of answers. By solving various problems, we simplify life. We’ push back against the force of entropy.
Precisely right! cries the minimalist. By decluttering — by limiting purchases of unnecessary possessions in the first place — this is how we fight entropy. This is how we create order. The minimalist’s mission is much broader than having a neat house — it’s to make sense of a complicated world.
Power From Training
It should be self-evident that to solve problems and simplify life, we need computational power. The more, the better. We need to be intellectually strong and fit.
Large language model generative AI applications like ChatGPT require something like the equivalent of 10,000 high end processors to operate.(4) It’s nice that we can access such powerful tools. But what about our brains? How do we maximize our own computational fitness?
Human brains differ from computer applications in being interconnected with human bodies. Our thoughts are linked inextricably with feelings and sensations. With stress. With fatigue. A large body of research finds a consistent association between mental performance and physical health. Writing in the 1960s, JFK made exactly this point, when he warned that moral courage requires physical health:
For physical fitness is not only one of the most important keys to a healthy body; it is the basis of dynamic and creative intellectual activity. The relationship between the soundness of the body and the activities of the mind is subtle and complex. Much is not yet understood. But we do know what the Greeks knew: that intelligence and skill can only function at the peak of their capacity when the body is healthy and strong; that hardy spirits and tough minds usually inhabit sound bodies
— John F. Kennedy, “The Soft American,” Sports Illustrated, December 26, 1960
There’s not much mystery to developing a healthy body. You could, for example, take up running. (Or some other sport.) The real question, in today’s world where people are subject to information overload — and constantly distracted, is how do you find time to train?
That’s easy. By focusing on what’s important and ignoring things that don’t matter. By learning to say “no.” For example, you could push back on advertising. You could stop buying junk. You could declutter your house and free up space in your mind.
What I’m saying is – if you want to be a minimalist, you’ve got to be fit. And the way to get fit is to go minimalist.
I see no alternative — do you?
Running the Long Path is available on Amazon!
(1) For a discussion of information destroyed in a simple arithmetic calculation, see https://www.scientificamerican.com/article/the-fundamental-physical-limits-of-computation/
(2) https://fs.blog/occams-razor/
(4) https://www.stylefactoryproductions.com/blog/chatgpt-statistics
The Long Brown Path 