Home » Burn the Map: AI Agents, Trust, and Why “Do AI” Fails w/ David Espindola

Uncategorized

Burn the Map: AI Agents, Trust, and Why “Do AI” Fails w/ David Espindola

In This Episode:

We talk to David Espindola about the part of the AI boom most companies keep trying to skip: the boring, unsexy work that makes AI real—problem definition, strategy, governance, and trust.

David’s written two books (The Exponential Era and Soulful) and spends his time thinking about what happens when exponential tech stops being a keynote and starts breaking things in production. We get into why “go do AI” is an executive-level failure mode, how to decide whether you’re a maker, shaper, or taker, and what it actually takes to build AI agents you can trust—especially when incentives (ads), privacy, and security start pulling the strings.

What We Cover:

  • Why “we must be an AI company” is a terrible mandate if you can’t name the problem you’re solving
  • The Makers / Shapers / Takers framework (and why most teams should aim for “shaper,” not “maker”)
  • Why AI ROI is so messy right now (and how bad benchmarking + vague goals make it impossible)
  • Agents and non-determinism: the uncomfortable truth that “it worked 26 times” means nothing
  • How David built Zena, his AI colleague—using RAG to ground the assistant in his knowledge, values, and working style
  • The trust-killers: ads, misaligned incentives, weak security, and “agents in prod” without guardrails

Guest Bio:

David Espindola is the author of Soulful and The Exponential Era, and the creator of Zena, an AI colleague built around a simple premise: if you can’t trust the system, it’s not helping—you’re just outsourcing risk. David works at the intersection of AI strategy, change management, and human–AI collaboration, with a practical bias toward governance, security, and using AI where it actually fits (not where it looks good in a board meeting).

Enjoy! This show was Brought to you by Wrench.ai

Follow Dan:
LinkedIn: https://www.linkedin.com/in/danbaird/
X: https://x.com/mrdanbaird

Follow David:

LinkedIn: https://www.linkedin.com/in/davidespindola/

Follow the Pod:
YouTube: https://www.youtube.com/@burnthemappodcast
Twitter/X: https://x.com/BurnTheMapPod
Instagram: https://www.instagram.com/burnthemappodcast/
TikTok: https://www.tiktok.com/@burnthemappodcast
BlueSky: https://bsky.app/profile/burnthemappodcast.bsky.social

Selected Links From This Episode:

People and Organizations Mentioned:

  • David Espindola
  • Dan Baird
  • Wrench.ai
  • McKinsey (makers / shapers / takers framework, referenced)
  • OpenAI
  • Anthropic

Show Notes & Timestamps:

  • 00:57 — Why The Exponential Era came first; the SPX methodology (Agile + design thinking + more)
  • 04:41 — Soulful: what AI is / isn’t; emotional intelligence, trust, and turbulence ahead
  • 07:51 — Biggest implementation mistake: “AI” as mandate vs starting with the problem
  • 08:xx — Makers / Shapers / Takers strategy framework
  • 12:26 — Abundance vs “trickle down”; why ROI is messy in early days
  • 16:43 — 50-year hindsight: today will look primitive
  • 19:46 — The myth that tech automatically reduces hours worked
  • 23:12 — Zena: training an AI colleague; trust as the core feature
  • 24:54 — Ads and incentives: when agents stop being trustworthy
  • 26:26 — How David trains Zena (RAG, memory improving)
  • 29:28 — Non-determinism: why testing AI isn’t like testing traditional software
  • 31:35 — When “agents” turn back into old-school software (APIs, reliability, cost)
  • 33:52 — Where to find David + Brainyus + Zena

New Episodes

Subscribe to our Newsletter!