Episode Description
Most conversations about AI ask "how do we control it?" Sharna Fabiano asks a more useful question: how do you lead AI so it doesn't end up leading you? In this episode, she talks with Elena Filatova about the mechanics behind how AI actually generates answers — and why understanding those mechanics is what keeps a human in the lead role, instead of quietly following whatever the model produces.
Elena Filatova spent seven years at Google in data infrastructure and business intelligence before going independent. She now runs BCB Code Labs, where she builds AI-driven applications and helps businesses automate with AI, and teaches AI literacy.
You'll learn:
- How one changed word can completely change an AI's response, and the mechanism behind it
- Why AI "hallucinations" aren't mistakes — they're a statistical feature, not a bug, and why that changes how much you should trust an answer
- A simple technique — assigning AI a "persona" — for sharper, more targeted answers
- Why asking AI to "make a list" is a more useful prompt than it sounds
- Where to draw the line between real human-AI collaboration and wasteful "AI slop"
Episode References
Attention is All You Need (2017), Google Research
Connect with Elena
*
Get a discounted coach session with Sharna with promo code PODCAST
https://www.sharnafabiano.com/coaching/
Order the book: Lead & Follow
https://www.amazon.com/Lead-Follow-Dance-Inspired-Teamwork/dp/1646632796/
Support the Show!
https://www.buzzsprout.com/1735834/support