We live in a moment of change regarding the idea of AI. Large language models are slightly better at impersonating thought than they were a year or five ago. The way we talk about robots and chatbots and panopticons is evolving. And yet, the actual digitized capitalist hellscape we live in continues to demonstrate every day, in every way, that robots are just playing stupid human games at 10x speed. The most innovative thing about technology is the errors it makes. A bot made to look at pictures thinks that hills and sheep are the same thing. If you ask “How many giraffes are in this picture?” it will always say one, because otherwise you wouldn’t be asking about giraffes. When we someday make an android that passes for human and stuff ChatGPT3000 inside, the effect will be that of someone given daily doses of ecstasy in utero who somehow survived to adulthood.
But we keep letting robots do things. Both humans and robots are going to learn a lot of things the hard way. We can only hope it’s going to be Tesla Fail hard and not Paperclip Maximizer hard.
Most of the time in this experimental year of 2023, it’s just silly, frustrating, or both.
I got an email this week that appears to legitimately be from Bytedance’s “publishing imprint.” Reading it, I wondered how much of it was written by a human. It certainly wasn’t aimed at me by a human, because a human would have taken one glance at my public existence and realized that I was the worst possible target. To be completely honest, I responded to it because I want to see if a human reads the responses, and if so, what their reaction might be. And I’m posting the exchange here because it contains some advice for writers. It might even be a good exercise in spotting red flags. I enumerated a few of them. How many do you see?


Edit to add: I received a response that I’m fairly confident came from a human. It says my concerns are valid, then makes no attempt to address or defend them. The overall vibe is “Whoops, sorry, wrong number.”

I think, from now on, I will respond to all such emails. Just to make a human think about what they’ve done and how well it’s going for them.