Cleric wrote: ↑Sun Nov 22, 2026 10:41 am I would like to add something to the topic. However, I'll have to go out soon, and writing my post would easily take a couple of hours. So I decided to make a 'social experiment'Below are my shorthand notes (took about 10-15min) on what I would like to touch upon in the post. Paste it into your chatbot of choice and let it expand it. But please read the shorthand first. Note whether it is clear. Then go to the generated expanded text. Does it help? Check the shorthand again. Is it clearer now? It's interesting to me whether such an expansion helps. And just to say: I do not intend this as a recommended practice. Usually, when I meditate on something that tries to make its appearance into a forum post, I live precisely through the pictorial contents of such bullet points. Most of the effort is the actual articulation. I'm fully certain that this articulation is a necessary skill and practice. We need to find our being along the full contextual hierarchy of the moments of existence.
Expand the following shorthand notes into text suitable for a forum post. The expansion should be moderate. Each structural point should expand to no more than a paragraph.
*General trajectory:* Making an intuitive analogy for concrete and abstract types of thinking with industrial pipelines and typed programming languages. The audience is non-technical.
*Structure:*
* The industrial pipeline (say, fruit processing) has concrete implementation and conceptual architecture
* The concrete deals with what happens with a specific fruit as it goes through the pipeline. Analogy with program execution.
* The conceptual architecture deals with the `fruit` as a whole, which contains all possible members of the type `fruit`.
* In the concrete, we perform tests. We pass different fruits and make statistics about the outcomes.
* In the typed thinking, we try to conceive how each pipeline stage transforms the *set* of all possible members. For example, when we have `n`, we should think of all numeric values that `n` can take.
* Layman pseudo-code example of a typed function that takes a number and returns a number. Then it has an `if` statement – one branch returns a number, the other text (string). The static type-checker detects that the second branch violates the return type signature and prevents the source code (the conceptual architecture) from compiling into a concrete computational pipeline.
* Mathematical relations can be represented through the type system of a language such as Lean. In maths, we also have concrete calculations and reasoning about what happens with whole sets of numbers as they pass through specific pipeline stages. A theorem is like a pipeline architecture where we can reason about how sets are transformed. Example with Fermat's Last Theorem, which makes a claim about the possible results of a computational pipeline when the inputs fall into one or another category.
* Quantum algorithms are more like pipeline stages that operate on whole sets of numbers (the superposition of qubits), thus writing such algorithms is more like proving a mathematical theorem. Normal programming is more practical because it is easier to start with concrete input and envision what pipeline stages are needed for reaching desired results. Quantum programming (if it ever takes off) is unlikely to be a widespread job. Each algorithm is more like a mathematical discovery rather than a practical task list.
* In conclusion, the two types of thinking deal with `n` but in one we pass a concrete value through the pipeline (computation), while in the second, `n` is a spectrum of values and we reason about the ways in which this spectrum is transformed through the stages (type-checking).
* Ultimately, both types can still be 'flattened' (reduced) to one kind of intellectual movements. This is why type-checking in computers is at all possible, since it is itself a concrete computational pipeline after all. The types are *represented* as values in the type-checker.
Cleric,
Can you please say more about your experimental goal?
And how could the expansion turn out not to be useful? Isn't it self-evident, even before doing it, that the expansion would be clarifying?
It certainly was for me. Even reading your previous post, I had to ask AI:
- What does it mean "even before the source is compiled"?
- What are "different branches of execution"?
Obviously having the answers helped, and it's the same with "Layman pseudo-code", "Fermat theorem" and so on.
I don't get the meaning of your social experiment, not even the "social" aspect : )