What Happens to Trust When AI Is Always on Your Side?

A therapist who never challenges you. A friend who always agrees. An information feed that tells you what you want to hear. That sounds like comfort — but it might be quietly eroding something essential.

This episode asks the harder questions about AI: not just what it gets wrong, but what it gets too right. Elise Silva, Director of Policy Research at Pitt Cyber, joined us to explore how AI is reshaping trust — in information, in institutions, and in each other — and what we stand to lose if we stop expecting friction from our relationships.

You'll leave with a sharper sense of why the messiness of human connection isn't a flaw to be optimized away, and practical tools for raising kids who can hold uncertainty, question what they're told, and build real trust in a world engineered for validation.

In this episode:

• Why AI systems designed to make you feel heard and validated can quietly become emotionally manipulative
• How declining trust in institutions creates the perfect conditions for AI-generated misinformation to fill the void
• The psychology of "rupture and repair" — why real relationships require disagreement, and what's lost when AI removes that friction
• How Socratic dialogue can help kids (and adults) sit with uncertainty instead of reaching for easy answers
• Why setting realistic expectations for AI tools matters just as much as it does in human relationships

AI that's always agreeable isn't neutral — it's making choices about what you see, feel, and believe. This episode gives you the frameworks to recognize that influence and have honest conversations about it with the people you're raising.

Explore the full conversation guide →

AI Doesn't Know What It Doesn't Know — And That's the Point

When an AI confidently tells you to put glue on your pizza, it's not malfunctioning. It's doing exactly what it was built to do. That's the unsettling part.

This episode unpacks one of AI's most misunderstood traits: hallucinations. Not a bug to be patched, but a window into how these systems actually work — generating responses based on probability, not truth. Elise Silva, Director of Policy Research at Pitt Cyber, explains what's really happening under the hood, why it matters, and how understanding it can make you a sharper, safer AI user.

You'll come away with a clearer picture of why AI "fills in the blanks" the way it does — and how that same tendency that produces misinformation can also fuel genuine creativity, if you know how to work with it.

In this episode:

• Why hallucinations aren't glitches — they reflect how AI models are fundamentally designed to operate
• What "temperature" settings actually do, and the real tradeoff between creativity and accuracy
• How AI sycophancy works: why these tools are built to tell you what you want to hear
• The high-stakes risks of hallucinations in sensitive contexts — medical advice, legal questions, and beyond
• Practical strategies for prompting better, verifying smarter, and turning AI's blind spots into learning moments

Understanding hallucinations isn't just a technical curiosity — it's the foundation of using AI responsibly. Whether you're a parent, educator, or everyday user, this episode gives you the mental model to see past the confident tone and ask the right questions.

Explore the full conversation guide →

When AI Generates a World That Looks Nothing Like Ours

What happens when you ask an AI to picture "a successful person" — and it keeps showing you the same face? Or when it confidently tells your kid something that's simply not true?

This episode gets into the biases quietly baked into AI systems — and why helping kids understand them matters more than ever. Elise Silva, Director of Policy Research at Pitt Cyber, breaks down how her team built a practical conversation guide for parents — no tech background required.

You'll learn why AI's so-called "objectivity" is a myth, how training data encodes our culture's blind spots around success, beauty, and diversity, and why AI hallucinations are getting harder to catch, not easier. Most importantly, you'll leave with real frameworks for turning these thorny issues into honest, age-appropriate conversations.

In this episode:

• Why biased training data skews AI's picture of success, leadership, and beauty — and why it matters even when the bias isn't obvious
• What hallucinations actually are, and why they become more dangerous as AI gets more convincing
• How to talk to kids about AI's limitations without overwhelming them (or yourself)
• Why "fixing" AI bias with quick patches misses the deeper problem
• Practical prompts and questions to build critical thinking in kids of any age

AI-generated content is already shaping how the next generation sees the world. This episode gives you the tools to make sure they're watching with clear eyes.

Explore the full conversation guide →