
Morning Thoughts
I was startled when I read the last text I wrote with AI assistance – what a Mount Everest of characters aka letters aka text. A deluge of words. The joke: when I look at my prompts, their length, I could have just written the texts myself right away. I am a great fan of Schulz von Thun and his small book on how to express oneself more briefly and concisely. I had also tried it once in my legal world, but both courts and colleagues are addicted to wordiness and monstrosities of text. Sure, with the advent of automatic dictation systems, I saw some colleagues in my mind’s eye standing by the window and babbling, and this deluge of words was captured in texts that landed on my desk either digitally or classically on paper. There I had to dissect them until content could be recognized beneath all the layers of text.
AI set out to bring clarity, structure. Admittedly, that actually works quite well. The ascetic among the systems, NotebookLM, adapts well to one’s own needs. It only takes the information you give it. The philanthropist Claude is good at discussing the big picture, and the librarian Perplexity provides the necessary sources, while ChatGPT functions as a Swiss Army knife. I thought this little world would save me work, but that is only true to an extent. You have to contain them and must never really forget who you are writing for.
I like to write on paper, with a fountain pen or even a pencil. This physicality of writing is an insurance for me that what I write also has meaning, because thinking something is one thing, capturing something on paper is another. This cannot be compared to the process of typing on a keyboard or dictating. Both are only a single keystroke away from insignificance, whereas I simply cannot tear up important papers, but have to send them through a shredder. And who would really write personal thoughts on a computer? On the other hand: a physical diary is more vulnerable if found than a file that I delete. Whereby with a diary, the question is always: Who am I actually writing for? Do I assume that someone will read it – am I pleased if someone reads it and even likes it? Am I vain if I rejoice when someone likes it? Where is privacy, where is the public sphere? The boundaries have shifted, and our AI language assistants can only affect us so emotionally because we are unsettled about what is private, what our circle is.
It was decades ago, I saw a broadcast about predator training on a black-and-white television. The tamer – he was famous back then, and training animals was still considered normal – showed three virtual circles around the lion. If he was outside the three circles, the lion didn’t care. At the third, outer circle, he became attentive; at the second, he straightened up; only to react at the inner circle. The tamer showed how he moved within this pattern to trigger reactions from the lion.
We must learn again where our three circles are – we have obviously forgotten.
The Dialogue with the AI