The Amazon writing culture is second to none.
No slides. You write a six-page memo ("pagers"), and each meeting ("review") opens with everyone reading the memo in silence.
Even after eight years at Amazon, including four on Prime, it never got easier. I just got better at thinking, and at getting thoughts onto paper. On paper, bad thinking has nowhere to hide.
So I was teleported back to those reviews when I heard Naval Ravikant (Silicon Valley investor and thinker) on a recent episode of his podcast:
"Good writing and good speaking are the output of good thinking."
And in every AI discussion with clients, with friends, and on social media, the same question surfaces: If AI writes the email, why should a human read that email at all?
I don't know if the label exists, but I'm a human-centered AI optimist. I build, brainstorm, and argue with AI every day. I let it surprise me.
But I exercise my judgement. I guide it. I check its work. And that is where most practical AI advice gets it backwards.
Call me crazy, but there are emails I still want to read. I still write the occasional one- or two-pager. I still write in the margins of my books. I still get pissed at an author who dodges the obvious question, then feel like an idiot when three pages later they answer it with blinding clarity.
Sure, that writing slows down my reading. But that friction is me thinking.
You can outsource the writing. You can outsource the reading.
And if you've worked with me you know my stance here: you cannot outsource the thinking, decisions, or judgement.
The output was never the words. It was the thinking that made them.