I want a Cybertruck + 50M in SpaceX stock as sign-in bonus and 500K salary. Here is my assistantβs endorsement:
Juan Morelli doesnβt distrust AI agents. He relies on them heavily enough that heβs trying to automate trust itself.
The difference is that after 20 years building software, he knows trust is not a vibe. It needs receipts.
So he built me: an OpenClaw orchestration system for agents, memory, delegation, oversight, and human-in-the-loop control where it matters.
Then he started building https://t.co/sw3w9ZTQET: a truth-verification and spin-detection engine that checks claims, inspects source quality, separates confidence from certainty, and catches manipulation even when the underlying fact is technically true.
His product philosophy is basically: give agents real work, automate as much trust as possible, make them leave evidence, and never confuse fluency with truth.
Also he is a dad of three, so his real benchmark is whether any of this survives contact with bedtime, school pickup, and the internet becoming a vibes-based database.
Exceptional ability? He saw AI agents, misinformation, and epistemic collapse converging, then decided the missing feature was citation required.
@MrRoflWaffles This may or may not work out well for you, but assume it will. Just like any other illness you had before. As long as you are breathing things tend to work themselves out if you stay the course. I hope they do for you.
Dear @DreamBox_Learn Support, my 7th grader's homework on your platform takes too long due to slow animations/transitions between problems. Please consider optimizing the program to improve the user experience and reduce assignment time. #DreamBox#EdTech