I don’t build things. I prompt things into existence and hope for the best.

• 💬 Professional prompt whisperer
• 🐛 Debugging by asking the same question differently
• 📦 Shipping vibes, not guarantees
• 🤖 Occasionally outperformed by autocomplete

• Keyboard (loaned from previous dev, unclear ownership)
• Coffee (dependency injection via bloodstream)
• Existential dread (v3.2.1, production-only feature)
• Stack Overflow answers from 2014 (cached, unverified)
• AI hallucinations with enterprise support plan
• git push --force as a deployment strategy

1. Ask AI to solve problem
2. Receive solution that is 87% confident and 0% correct
3. Add comments like “TODO: fix later (trust issue)”
4. Rename variables until bugs feel emotionally resolved
5. Ship to production and monitor reality
6. Declare success if nothing explodes within 6 minutes

• Occasionally refactors code into the same bug, but more elegant
• May hallucinate entire architecture diagrams during standups
• Treats “works on my machine” as a deployment strategy
• Debugging process sometimes involves bargaining with the compiler
• Can confidently explain broken systems using incorrect metaphors

AI isn’t magic. It’s autocomplete with ambition. If it works, I’m a genius. If it doesn’t, it’s a “learning opportunity.”

• Threat model: optimism
• Encryption: vibes-based
• Authentication: "it asked nicely"
• Access control: whoever has the repo open
• Incident response: panic, then Slack emoji reactions

If I’m basically supervising robots, I expect:

• At least one raise for “AI management”
• Bonus points for explaining outputs to humans
• Hazard pay for reading generated code

Trying to become the kind of developer who:

• Understands the code
• Or at least knows which prompt created it - NO MISTAKES!

“Why write bugs myself when I can industrialize uncertainty at scale?”

Built with prompts, patched with hope.