How do you get an honest picture of your whole life with AI in one night?

One night in August 2024 I spoke everything going on in my life into a microphone and handed it to a model. It became a report of thirteen pages. The value came afterwards: print it, test every area for whether it brings joy, and break what survives into three movements. Delegate, automate, systematise. It was 5:19 in the morning. I was sitting in the coworking space, after telling myself the evening before that I would not leave until the video was done. Around 272 videos lay on the channel, and still nothing new had appeared for months. That night I looked at my own overload from the outside. A life is hard to sort inside your head. On paper it works. Once everything sits in one place you see the duplicates, the dead strands and the spots where too much runs at once. The document is not yet a result. It is what makes a decision possible.

Why a document about your own life helps

A life is hard to sort inside your head. On paper it works. Once everything sits in one place you see the duplicates, the dead strands and the spots where too much runs at once. The document is not yet a result. It is what makes a decision possible.

Let us be honest, life is very easy, it is just really hard to do what you already know.

In the video that sentence sits above sleep, food and movement. It holds for work too. You usually know what should be done. What is missing is the overview that forces the decision.

The sequence of one night

  • Speak instead of type: press the microphone in the chat and tell it everything that is running. Projects, dates, people, open questions, the uncomfortable ones first. Speaking is faster and more honest than writing.
  • Check the raw text: read it once, correct the words that were misheard, and add what you left out while speaking.
  • Cross-check with several models: send the same text through two or three of them. Each finds other gaps, and those differences are the yield.
  • Condense: build one document from the versions, cleanly structured and in parts if needed, because very long texts rarely come out clean in one pass.
  • Print it: only on paper do you notice how much is actually there.

The filter that turns a list into a decision

Marie Kondo tidies flats like this: pull everything out of every corner, put it in one pile and take each item into your hands. The question is whether it brings joy. Whatever does not feel good goes. That night I asked it about projects instead of objects, area by area, with the printout in front of me.

For what survives the filter there are three ways.

I can delegate, I can automate and I can systematise.

Delegating means a person takes over. Automating means a machine takes over. Systematising is usually a mixture of both plus a small remainder that stays with me. A fourth way came up in passing back then.

The third thing with a lot of leverage is content, because it lets me talk to more than one person at a time.

This garden is exactly that. A thought written down cleanly keeps working while I sleep. What I mean by leverage is on the page about it.

What is left of it today

Two years later more has survived than the document. The voice dump has become a second brain, an ordered collection of notes, decisions and reasons that a person and a machine read equally well. A report goes stale in four weeks. A collection that every conversation flows back into grows.

The second change concerns direction. Back then I talked and the model listened. Today I let it ask first. The prompt is one sentence: based on what you know about me and my goals, which information can I still give you so that you help me better. The technique is called reverse prompting, and in twenty minutes it surfaces what took me a whole night. The sequence is in Should the AI interview you before you prompt it?

The third change is the least visible. I no longer copy text by hand through several windows. The agent reads what is written about me itself, because it has read access to the right folders. The context now comes from access instead of from the clipboard. Why that carries further than wording is in Why does access beat the prompt?

What stayed the same is the order. Throw out first, then delegate, then automate. Anyone who automates a task they could have deleted has kept it forever. There is a note for that, What do you delete before you automate? How far an assistant may go with you is in Which rung of the Jarvis ladder are you really on?

At the end of the video I say I am curious to look back on it in three, five or ten years. Two have passed. A good share of those projects no longer runs. That is exactly what the filter was for.

From the video

Wie ich mein Leben in einer Nacht mit KI organisiert habe, YouTube, 8 August 2024, 34 minutes 43 seconds.

  • Recorded shortly after five in the morning in the coworking space, after a promise the evening before not to leave without a video.
  • Result of the night: a report of thirteen pages about my own life.
  • Step one is the voice dump into the model, not typing.
  • Step two is the cross-check: the same text through several models, because each finds other gaps.
  • The filter is the question about joy, applied to projects instead of objects.
  • The levers afterwards: delegate, automate, systematise, and content as the fourth one.

This note keeps growing

2026-09-03: Planted from the video "Wie ich mein Leben in einer Nacht mit KI organisiert habe", 2024.

ZukunftBilden GmbH · Salzburg · +43 681 81655313 · office@zukunftbilden.eu