Doing it yourself (or not)

Recently trying to do it all at once and becoming ChatGuiltp, which is the feeling of unease you get when using a Large Language Model to automate something you really should be doing yourself. Turns out, it isn't a good substitute for actually living and using your biological brain to do things, since sensing, dreaming, executing, communicating et al. are still steps above calculating (and we may find ourselves quite soon astonished that we ever thought it possible otherwise, but similarly we can get into situations where the automation is just too irresistible to avoid.)

I have to recognize that it's a tremendous tool: mere thoughts in my mind become fully fledged pipelines quite rapidly, and where previous blocks would have stopped me for weeks, possibly for ever, I can now turn to that dark grey page and trick myself into actually engaging with a project by formulating an attempt at describing what I need to an impersonal online tool that pretends to be an entity.

But the output is a little erratic which seems at odds to me with what automation was envisioned for: a set of rules, a consistent, repeatable standard that could "scale up", a recent term, behind which lie the paradigms of Taylorism and externalities which still bedevil us.

So the real show of intelligence and capability to see things get done might come from talking to others - in interiority, you're not worth much more than an AI, despite how brilliant it sounds in your head.

And we have our word writer for all the rest.