AI Made Engineers Faster. Leadership Fell Behind

Ep 17: AI Made Engineers Faster. Leadership Fell Behind

AI changed how fast a team can produce software. It did not answer the harder questions. What should you build? Who needs it? How will you know it worked? When should you stop?

Oscar Gallo and Matt Wozniak discuss those questions with Mike Lyons and Greg Pfister from KaiRise. Mike and Greg have worked across software engineering, government programs, consulting, Agile coaching, and product education.

Shipping on time can still be failure

Mike shares the project that changed his career. His team delivered a $4.3 million voter registration system on time, on budget, and against the written requirements. Users did not want it. The team delivered the requested output and missed the needed outcome.

That distinction matters more after AI. A team can now build the wrong feature faster and produce more of it before a customer sees it.

The slow part moved

Greg argues that engineering was often only one part of a much longer delivery process. Requirements, approvals, testing, budgeting, hiring, sales, and product decisions still took time. AI made the engineering step faster, which made the delays elsewhere easier to see.

Mike's challenge to leaders is direct. If a team can build tonight, leadership cannot wait six months to decide what comes next. Planning horizons must shrink. Budgets need more frequent review. Product direction must become clearer.

Agile principles without Agile theater

Greg still values fast feedback, transparency, customer contact, and frequent delivery. He cares less about whether every team follows the same sprint length or ceremony.

Mike calls this lowercase-a agility. The principles remain useful. Formal roles and fixed frameworks need to prove that they help a team respond faster and deliver better outcomes.

Stop measuring activity as value

Story points, lines of code, and token usage can show activity. They cannot prove that a customer problem was solved.

The group returns to one question throughout the episode: should we build it?

Fast software production creates an endless list of possible features. Product judgment determines which ones deserve time and money. Teams need customer feedback before context and sunk costs make a weak idea feel permanent.

What AI cannot solve

AI can suggest options. It cannot replace the human work of building trust, resolving conflict, judging tradeoffs, showing empathy, and choosing a direction.

Engineering and product teams need a shared understanding of the customer, the business goal, and the proof that the work created value. As software gets cheaper to produce, those skills become more important.

Guests

• Mike Lyons, KaiRise
• Greg Pfister, KaiRise
• KaiRise: kairise.com
• Listener discount: Use code human for 20% off all courses.

Your hosts

• Oscar Gallo: AI Engineer and entrepreneur. I live in the intersection of engineering and businesses.
• Matt Wozniak: Serial Builder and relentless executor. I come from the lens of what works and what doesn't.

Listen now

• Watch on YouTube
• Listen on Spotify

What became the slowest part of your company after AI made engineering faster?