Alignment Drift
I am noticing a trend online. More of the writers and thought leaders that I follow are telling the world that they are not using AI, and that they did, in fact, think all of their own thoughts.
From Tom MacWright:
I do not use LLMs or AI tools to write, research, or publish anything on this website. Photos and illustrations are my own work unless otherwise noted.
From Seth Godin:
Everything on these pages was written by Seth Godin, not a staffer or an AI.
While I'm not quite as bearish on AI as some of my peers, I find myself drifting closer and closer to their position.
"When you stare into the abyss, the abyss stares back at you."
Friedrich Nietzche
That abyss quote from Nietzche is something I think about a lot. And I think you and the abyss are both doing more than just staring at each other.
When I interact with generative AI, like ChatGPT or Claude, I often find the AI is very opinionated in how it frames its response. But the framing is done in a subtle way, which, if you're not careful, you'll not even notice. And framing is a very effective persuasion technique.
I think I'm noticing what some other people already knew: AI is not a neutral tool or agent. It's opinionated. I believe more and more that AI is becoming a medium for thinking, working, and producing things. If people are not careful, the medium will become the message.