For a long time the comforting story about AI went like this: sure, it can crunch numbers and play chess, but it will never do the human things, the creative things, the intuitive things. That was the line, repeated confidently, for years. I want to gently point out that the line keeps moving, and it is moving in a direction that deserves more honesty than it usually gets.
The goalposts have wheels
The pattern is almost funny once you see it. AI will never beat a grandmaster, until it did. It will never write anything worth reading, until it wrote things people could not reliably distinguish from human work. It will never really code, until it started shipping code that professionals use daily. Every time, the response was not "I was wrong." It was to quietly move the definition of "real" intelligence to whatever the machines could not do yet.
What actually changed
The things we treated as uniquely, safely human, nuanced writing, visual creativity, reading a subtle piece of code and grasping the intent, are exactly where the recent models improved most. Not perfectly. Not everywhere. But enough that "AI cannot do the creative stuff" has gone from an obvious truth to a claim you have to carefully qualify. That is a big move, and pretending it did not happen is its own kind of denial.
Every "AI will never" has a shelf life. We just keep buying more of them because the alternative is uncomfortable.
Why we keep doing it
The goalpost-moving is not stupidity. It is self-defense. If the machine can do the thing we thought made us special, we would rather redraw the boundary than sit with the feeling. I understand the impulse. I feel it too. But a skeptic worth the name has to be skeptical of the comforting story as well as the hype, and "it will never do the human stuff" has become a very comforting story that keeps being wrong.
Where I actually land
Not at doom, and not at worship. The machines are getting good at things we were promised they never would, and they are still bad at plenty. Both are true. The honest posture is to stop treating any particular human ability as a permanent moat and start asking a better question: given that the line keeps moving, what do we actually want to keep doing ourselves, not because a machine cannot, but because we want to. That is a harder question, and a more useful one.