Whether ChatGPT, Claude, or whatever AI-assistant you maybe using, do you think of it as an actual person in your computer?
I do. Sort of.
Clearly, there is no little man in my machine. I know this both intellectually and emotionally. However, it helps me to think as if there is.
It helps me in two ways:
- The LLM “is like” a human in that it “understands” what I say, and how I say it. It helps me be better at giving specific instructions.
- The Agent can control my computer, has access to the internet, and can try things like me. Thinking of it like a person helps me be better at thinking about myriad solutions to the problem rather than just a specific set of instructions for a solution that I’ve devised already.
The guy in my computer is a mental model. Not a real one.
Maybe it’s creepy. Maybe it’s childish. But it certainly helps me.
Discover more from johnmaconline
Subscribe to get the latest posts sent to your email.