AI and the Future of Work: Artificial Intelligence in the Workplace, Business, Ethics, HR, and IT for AI Enthusiasts, Leaders and Academics

400: Wade Foster, Zapier CEO: What Happens When Every Knowledge Worker Gets a Coding Engine

Dan Turchin Season 7 Episode 400

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 40:10

Send us Fan Mail

Wade Foster is the CEO and co-founder of Zapier, the workflow automation and AI orchestration platform used by more than 4 million people to connect over 7,000 work apps. Since launching in the Y Combinator Summer 2012 batch, Zapier has automated more than 81 billion tasks, users have built more than 25 million Zaps, and the company has bootstrapped its way to a valuation north of $5 billion.

He holds degrees in industrial engineering and business administration from the University of Missouri, and has spent more than a decade on a single problem: making the tools people work in every day work for them instead of against them. He is also one of the show's rare repeat guests. His first conversation with Dan was in January 2024, back when coding agents mostly did not work.

In this episode, Wade draws on running the automation layer that sits between thousands of enterprise tools to argue that the frontier model leap of late 2025 handed every knowledge worker an engineer, and that the companies still waiting on returns are the ones that never changed how the work itself gets done.

In this conversation, we discuss:

  • Why the late 2025 model leap means every knowledge worker now has an engineer, and what that changes about ordinary work.
  • Where deterministic workflows still beat AI at work, and why the most sophisticated teams refuse to choose between the two.
  • What happens when white collar workers start pasting API keys into files, and the third option most security leaders are missing.
  • Why Wade says culture, not tooling, is the real constraint, and what leaders get wrong when the board demands AI first.
  • Why individuals report real AI gains while their organizations report none, and what has to be rethought before that gap closes.
  • How Zapier built its AI Fluency Framework from its own teams, and why version one was already dated the day it shipped.

Explore this conversation:

00:00 Welcome to Episode 400

02:01 AI Fun Fact: Is Agentic AI the End of SaaS Tools?

04:35  Introducing Wade Foster: How Zapier Bootstrapped to 81 Billion Automated Tasks

06:01 From Coding Agents That Kind of Worked to an Engineer for Every Knowledge Worker

07:53 What Zapier Does Today: Deterministic Workflows Meet Non-Deterministic AI

11:12 Hybrid Workflows: Build It With AI, Run It Like a Program

13:30 MCP and the Orchestration Layer: API Keys, Agent Harnesses, and the Daily Recap

20:23 The Third Option for CISOs: Governing AI Agents Without Saying No

24:00 Foot Guns and Default Settings: The Product Judgment Behind Agent Permissions

27:03 Accountability When Software Is No Longer Deterministic: Evals and Spec Adherence

28:34 AI Strategy Office Hours: Why Individual AI Gains Are Not Becoming Institutional ROI

33:03 The Zapier AI Fluency Framework: Why Version One Was Dated on Arrival

36:15 Looking Ahead to 2028: The Hive Mind Company and a New Management Playbook


Resources: