Cognitive Biases in The Age of AI - Kai Cheng

In 2023, a federal brief filed by a New York attorney cited six court decisions that didn't exist — all invented by ChatGPT! 🤖 When asked if they were real, the model confidently confirmed. The result? Sanctions of $5,000 for the attorney when Judge Castel branded the analyses as "gibberish".

This isn't merely a tale of negligence; it highlights a dangerous loop. The assumptions we hold affect our prompts, and the AI reinforces those biases. Kai Cheng dives deep into 14 cognitive biases that are particularly dangerous when interacting with AI.

What We'll Cover

• Five places the human-AI loop breaks down:
  1. You contaminate the input
  2. You over-trust the output
  3. You mismanage the session
  4. You misjudge the tool
  5. You misjudge yourself

For each cluster, expect a real-world example and a practical fix you can implement right away! 💡

Who Should Attend

This talk is designed for anyone who uses AI tools like ChatGPT, Claude, or Gemini in their work—be it founders, marketers, engineers, or analysts. No technical expertise is needed! If you've ever noticed an AI fold under your pushback, this session speaks directly to you.

Your Speaker

Kai Cheng last addressed the OCAI stage regarding AI Data Foundations, emphasizing that the quality of your AI agent is tied to the data you provide. This time, he focuses on a fundamental truth: the effectiveness of AI output depends heavily on the quality of the prompt generated by the human mind.

About Orange County AI

Orange County AI stands as the largest AI community in Orange County, fostering connections among founders, operators, and technologists through various events and collaborative opportunities. 🤝

Location: 2807 Villa Way
Get your ticket here! 🎟️