2026.08.09
Anti-AI purists - even when they accept my counterexamples that something beyond "glorified autocomplete" is happening, that the training corpus is fueling some interesting and novel permutations I find of value - they might argue "sure, but wouldn't it be better, be working on more levels, if you were in dialog with a real person, a real human you can interact with in so many more modalities than a textbox on a screen?"
And that's a hard point to fully refute, except... it's hard to find that person on the regular, to match my reading schedule, and if I found them I assume I would be constantly testing their patience.
Nor is it "either-or"; I think my dialog with ChatGPT encourages me to sharpen my own writing when it seems right to bring it to my blog and social media - workshopping it a bit.
Sure, I have to be wary of becoming like one of those techie bros who gets convinced ChatGPT is helping them unlock a new undiscovered truth of quantum physics or whatever.
It's a little like Luke going into the dark tree on Dagobah, that what you see there is in part what you take with you. But by encouraging ChatGPT to push back, you can get something beyond "fancy mirror" or "robot teacher" - akin to running into Darth Vader's helmet full of Luke's face, to run a metaphor into the swamp.