Overview
AI Decoded is a two-session AI literacy program I designed and taught at the YMCA of Metropolitan Chicago in July 2026, with sponsorship from AT&T. The sessions ran on July 11 and July 18 for small groups of community members, three to five people each, which left room for every question.
The program explains, in plain language, how AI systems produce their answers, where they go wrong, and how to check an answer before relying on it. One section covers bias: AI systems learn from human writing, flaws included, so our stereotypes and blind spots travel with them. “The computer said so” does not make an answer neutral. Participants practice asking whose view might be missing from an answer and treat that as one more reason to verify anything that matters.
What I did
- Designed the two-session curriculum and the slide deck.
- Taught both sessions on July 11 and July 18, 2026.
- Secured AT&T sponsorship for the program through the YMCA.
- Covered how AI systems learn from human writing and inherit its biases.
- Taught participants a simple routine for verifying AI output before acting on it.
Results
- Two sessions delivered, with three to five participants in each.
- Small-group format meant each participant left with their own questions answered.
Why I built it
People are being asked to trust AI answers at work, at the doctor, and at the bank without being told how those answers are produced. The aim of the program is that participants leave knowing where an answer comes from and how to check one before they rely on it.