It’s almost 23 o’clock and past bedtime, but I quickly want to get this off my chest. Since a long time I wanted to create a table of contents for my blog, but held back because of the lack of excerpts. I didn’t write any. So tonight I was wondering if there’s some machine who could do that for me. And mother of god, there is.
I stumbled upon the website openai.com which provides access to an Artificial Intelligence. It can do all sorts of things, but it can also read books and write summaries. So I fed it a couple of my blog posts.
I’m still not sure whether I should be amazed, flattered or thoroughly scared. I don’t know how to feel about the current state of Artificial Intelligence. But what I do feel is… that it can write great summaries. And that it somehow seems to understand my writing better than I do myself. These summaries are really helpful for me as feedback, to understand what it is that I am writing.
For example, my recent text, “A taste of freedom – Trauma healing in Somatic Education.” So how it works is this: I use the web-interface for convenience, and there I tell the Artificial Intelligence in plain English what I want, and then it answers me. Here’s how this looks like:
Version 1: I wrote, “Summarize the following text.” And then added my blog post. The result was this:
The text discusses the work of famous trauma-experts Bessel van der Kolk and Peter A. Levine, and how they fail to mention the role of compulsory schooling in causing trauma. The text also discusses the work of psycholinguist Frank Smith, who is not afraid to look at the origins of trauma in compulsory schooling.
Version 2: I wrote, “Write an angry sounding summary of this text.” And then added my blog post. The result was this:
This text is a long-winded diatribe against famous trauma therapists who fail to mention the role of compulsory schooling in causing trauma. The author bemoans the fact that these experts are not brave enough to speak out against the machine.
Version 3: I wrote, “Summarize this text in one word.” And then added my blog post. The result was this:
I kid you not, this is what the Artificial Intelligence wrote back to me. I’m tempted to change the phrase “The result was this” into “The answer was this.” As if it was a real person answering, someone with a heart and compassion and humour and all that, not a machine.