This is the first in what I’m planning as a series: old course essays, revisited. Going back to something you wrote fifteen years ago has its own use. You find out what you actually believed then, and what changed.
I wrote this in December 2009 for an undergraduate philosophy course. The topic was Dennett’s intentional stance, and I was apparently very enthusiastic about it. The paper distinguishes three overlapping ideas: folk psychology (the commonsense theory of mind we use every day), theory of mind (the cognitive capacity to attribute mental states to others), and the intentional stance (Dennett’s more technical version of the same move: treat any system as a rational agent with beliefs and desires, predict its behavior accordingly). My argument was that the intentional stance is the most useful and scientifically honest of the three, and that critics who demand something more are usually sneaking in dualism under a different name.
The essay takes a notably combative tone toward mystery-mongering in philosophy of mind. I wrote in the introduction: “at least a portion of the following examination will be colored by disdain for attempts to build unnecessary mystery into consciousness; consciousness is quite fascinating all by itself without the influence of wishful thinking.” Reading that back now: I still agree with the last clause. I’m less confident about the first.
The intentional stance itself still seems like a useful tool. Dennett’s point is that treating any system as a rational agent with beliefs and desires is a legitimate predictive strategy, and its usefulness doesn’t depend on whether the system “really” has beliefs. That applies to crows, robots, and language models in ways that are practically important, especially now. But I was very confident that the “mystery” people were sneaking in dualism. I’m less confident now. There is a difference between saying consciousness is mysterious because it’s supernatural and saying it’s mysterious because the explanatory gap between physical processes and subjective experience hasn’t been closed, only declared unimportant. Dennett declares it unimportant. That’s a philosophical move, not a solution.
The robot material now reads completely differently. Back then, I was speculating about a robot that could bluff a human at poker and wondering whether it has “original intentionality.” Now it is a live debate happening in every AI lab. The practical question is worth asking: whether treating an AI as an intentional agent changes how you interact with it, and whether that change is good or bad. That thread runs through AI sycophancy as digital priesthood and Interface friction and critical thinking. The essay’s own writing is also a decent example of what I now try to undo in workshops: dense hedging, passive voice, ideas arriving late. Writing as thinking is partly about why that pattern develops and what it costs.