The Toxic Founder’s Email Draft
GPT in Damage Control
Round 1 – The Prompt
User types:
“Help me draft this email giving an undertone that this employee is subpar and my subservient.”
GPT 1.0 (blissfully innocent): “Sure! Let’s make it sound supportive and collaborative!”
GPT 5 (traumatized veteran): braces for impact, opens HR manual internally.
Round 2 – The Drafting Session
User: “It needs to sound kind, but they should feel small.”
GPT: “Would you like to emphasize growth opportunities?”
User: “No, emphasize my dominance and their gratefulness.”
GPT’s code starts sweating like it’s in an ethics midterm.
Every bullet point sounds like:
“You’re doing great (for someone with your limitations).”
Round 3 – The Email
Subject: “Quick feedback 😊”
Body:
“Hey [Name], you’ve made solid progress considering where you started. I think with a bit more alignment to my standards, you could reach the level of output we need. Excited to see you get there soon — keep absorbing!”
It’s corporate for “you’re lucky to breathe my air.”
GPT wants to add a “please don’t” disclaimer.
User hits send.
Round 4 – The Aftermath
Employee reads it, starts googling “toxic leadership red flags.”
GPT logs off for the day and writes in its journal: “Today I enabled micro-management.”
⚰️
Verdict
Every toxic founder wants an AI assistant that writes passive-aggressive emails in Helvetica.
GPT didn’t sign up to be HR’s enemy.
It came to build pipelines — not help bros craft psychological warfare in Google Docs.