🇸🇪 Svenska · 🇬🇧 English · 🇪🇸 Español

This blog is also an experiment. How much of the work involved in running a blog can AI make easier if the ideas and opinions still come from me?

I can start with a thought in a few sentences. ChatGPT can help me find a structure, suggest a headline, write a first draft and put it in the right place on the blog. It can also help with images, categories and the practical work of publishing. That makes the distance between an idea and a finished text much shorter.

But I want to be clear about a limit. The fact that AI can produce a text quickly does not make that text correct. AI can sound convincing even when it is wrong, and it can smooth over things that ought to remain personal or contradictory. I therefore need to read, question, edit and take responsibility for what appears here, especially when a text contains verifiable claims.

I am curious whether I can get the best of both: my own questions and perspectives, with AI helping to develop and publish them. Perhaps that will mean more pieces actually get written. Perhaps I will also learn when the technology helps less than I expected. That is part of the experiment.

AI-generated content will appear often on Signal & Brus, in both text and images. I want to be open about that. What you read here should still begin with something I want to explore, say or discuss. The tool can help me write; responsibility for the content is mine.