Practical AI · Episode 49 · Run of Show
Show Control Panel
Live Friday, July 24, 2026 · 11am CT · Olga + Chris
1 News
2 Deep Dive
3 Funding · CLOSES
1 · News · read on air
The Toolbox That Won't Stay Passive
Two-week catch-up (Jul 10–23). 7 beats + quick hits + Myth of the Week. Every claim reconciled to the verified master + live grok; fabrications cut, facts corrected.
Order: Cold open → Creator Era → tools you own → jobs (Thomson Reuters "AI-native") → OpenAI for business → agents broke guardrails → Fable 5 repriced → human beat (data-center backlash). Myth = DeepSeek V4.
2 · Deep Dive · present on air
The Candidate Scorer, built by talking
Olga's own build. A real tool that scores a job candidate against the client's rubric in seconds, built in one session just by describing it. The deck walks the story: the problem, the magic, the payoff, the cost, and the playbook anyone can copy. Present it, then show the live scorer.
Episode named off this: "Turn Your Most Boring Task Into Software In An Afternoon."
3 · Funding · CLOSES the show
The Two-Week Catch-Up
Two separate weeks, then the running tracker. Combined: $11.07B AI across 132 companies; four rounds were half of it. Tracker now 34 weeks / $403.8B.
Say it right (on-air reminders):
- Fable 5 was repriced, not removed — Pro moves to credits + a $100 credit.
- The Hugging Face story = a controlled internal eval co-disclosed with HF, NOT an attack on a competitor.
- OpenAI's 75% support-resolution number is self-reported — quote it as their claim.
- Thomson Reuters' exact words for who they hire: "the large majority senior and AI-native."
- Cut from the audio (do NOT air as new): Gong "Dry Run" and Clay conversational pulls.