Tell Me If Anything Was Ever Done

“Tell me if anything was ever done.”

The line appears as note 1365 in Jean Paul Richter’s translation of Leonardo da Vinci’s notebooks. The next note begins another fragment: “Tell me if ever I did a thing which me ….”

They are striking words. They are also fragments. They do not tell us when Leonardo felt discouraged, whether he was judging his life’s work, or why he left a particular project unfinished. We should resist turning a few surviving lines into a complete account of his private thoughts.

The question can still do useful work for us. What counts as having done something?

An Unfinished Work Can Outlive Its Purpose

Leonardo’s Adoration of the Magi offers a concrete example. The Uffizi’s 2018 exhibition account describes a commission from 1481 that he left unfinished when he departed for Milan. Another painting eventually filled the intended role.

The surviving work still gives viewers something to study. That does not mean the original commission was completed. It means a work can fail to reach one intended destination and remain valuable in another way.

That distinction helps us avoid two mistakes in our own work. We can dismiss an unfinished idea before anyone else has a chance to learn from it. Or we can call a promising sketch complete because producing it felt like an achievement.

Both errors become easier when the idea is exciting. A drawing can reveal a possibility. A working device has to do something. The evidence appropriate to one is not sufficient for the other.

AI Makes the Drafting Question Harder

AI tools make it easier to generate another outline, another paragraph, or another version of a program. That can help us get past a blank page. It can also give us more material to sort through before anything reaches another person.

Consider a proposed internal guide. Generating ten pages is an activity. Giving a colleague a guide they can use requires something more: a clear audience, accurate instructions, a usable example, and a way to identify what remains uncertain. The length of the draft tells us little about those qualities.

The same is true of a software prototype. It may be ready to demonstrate an idea while still needing substantial work before anyone should depend on it. Naming that purpose makes the next step clearer. A prototype should be judged by the question it helps answer; an operational tool needs to meet the obligations of actual use.

Define a Useful Next Version

Choose one unfinished idea and name the person it could help. Then define the smallest version that would let that person respond to something concrete.

For an essay, that might mean one complete argument with its supporting sources. For a guide, it might mean one task explained well enough to try. For a prototype, it might mean one demonstrated behavior with its limitations plainly stated.

Check the claims and repair the parts that would mislead the reader. Share the work when it is fit for that stated purpose. Ask what was unclear or missing, and use the response to decide what to improve. Some work will need more checking before it can be shared; calling it a draft does not remove that obligation.

Leonardo’s unfinished work does not give us a rule to publish everything immediately. It gives us a more interesting tension: completion matters, and unfinished material can matter too.

The question is what another person can do with what we leave them. A useful version gives the next person somewhere to begin.