From jailbreak justice to grief-bot girlfriends, AI is seeping into places we never expected—and in this latest episode of Lunchtime BABLing, the BABL AI team takes a break from regulation and risk frameworks to explore the more wild side of AI adoption.
BABL AI CEO and AI Ethicist Dr. Shea Brown is joined by COO Jeffery Recker and CSO Bryan Ilg for a fast-paced, deeply human, and often hilarious tour of viral AI use cases pulled straight from the internet. What starts as a lighthearted list of strange AI behaviors quickly turns into a meaningful conversation about bias, loss, connection, and the blurry boundaries between human and machine.
Not Your Usual Compliance Episode
This installment moves away from the EU AI Act and ISO 42001 playbooks to focus on real-life, emotionally charged encounters with generative AI. Think: a ChatGPT jailbreak that helps someone escape jail time, digital avatars testifying in court, and people building chatbots to talk to their dead best friends.
For Shea, Jeffery, and Bryan, these use cases aren’t just bizarre—they’re a window into how AI is changing human behavior in real-time, raising questions about emotional well-being, cognitive bias, and even the future of relationships and education.
Is Talking to a Dead Friend’s AI Twin Therapy—or Trouble?
One of the most surprising segments explores the growing use of AI in grief and mental health support. From texting simulated loved ones to replacing human therapists with chatbots, the group unpacks the ethical dilemmas behind these tools—and whether they offer healing or harmful detachment.
As Bryan puts it, “There are stages of grief for a reason.” Shea adds that while AI may have a role in providing behavioral advice, it can’t replicate human empathy. Still, the proliferation of these tools raises a crucial point: the line between assistive and exploitative is razor-thin.
Education Enters the ChatGPT Arms Race
In another highlight, the team discusses a professor’s decision to start using ChatGPT to grade student essays—because students were already using it to write them. The discussion quickly broadens to the future of education in the AI era: Should we ban these tools, or build pedagogy around them?
The takeaway is clear: it’s not about banning AI, but about rethinking what learning, expression, and understanding mean in a world where knowledge is cheap and language can be outsourced.
Love, Bias, and AI-Generated Breakup Analysis
The episode wraps with a double-header: someone who fed ChatGPT every interaction with a crush to decode signals of interest, and another who asked it to analyze a recorded argument with an ex for signs of gaslighting.
Amusing? Yes. But also a warning about confirmation bias and the tendency to outsource emotional labor to machines that were trained to please, not to challenge. As Jeffery notes, “If you’re relying on AI to manage your relationship, you might already have your answer.”
Why This Episode Matters
As AI systems grow more embedded in our daily lives, understanding their social impact becomes just as important as understanding their technical risks. This episode of Lunchtime BABLing explores that cultural shift with candor, humor, and ethical insight—and reminds us that AI governance is about more than compliance. It’s about the future of human agency.
Where to Find Episodes
Lunchtime BABLing can be found on YouTube, Simplecast, and all major podcast streaming platforms.
Need Help?
Looking to explore responsible AI beyond the headlines? Visit BABL AI’s website for more resources on AI risk, algorithmic audits, and compliance.
