“If you’re just looking for a compliment machine? You’re in the wrong damn tab.”

That’s what LizaGPT said when one of my business partners, Justin Parnell, asked me to put it to the test.

Justin is a master at building AI agents and automation. After seeing how my digital twin works, he was curious: “Love how the digital twins are showing people how much the regular chatbots are prompted to be agreeable with their discourse. Let me know what Liza GPT says!”

LizaGPT did not disappoint (see below).

Most AI has a positivity bias. It’s trained to be agreeable out of the box because that keeps people engaged. But that’s not how real work is done. You don’t hire a strategist to nod along. You hire them to poke holes, stress-test your ideas, and help you build something better.

That’s why I trained LizaGPT to push back. To ask tough questions. To call out fluff when it sees it.

Unless you want to feel like (but aren’t) a rockstar all the time, train your digital twin to resist being a yes-bot!

Have a great weekend everyone!