165: "What Caused AI to Skyrocket in 2026?" ft. Justin Coats

Erik and Justin dig into what it actually means for AI to “become real,” arguing that consumer usage does not equal workplace adoption. The conversation lands on a “pause moment” where organizations are finally forced to address governance, security, policy, and measurement because the tools are now powerful enough to create real operational risk and real operational leverage.

🧭 Conversation Highlights

• Justin distinguishes massive user counts from meaningful adoption, emphasizing that most people use AI on free tiers or inside existing apps without realizing it is AI.
• Both agree the real shift is B2B and workplace implementation, where adoption breaks down into training, governance, governance-adjacent policy, and data access safeguards.
• They compare possible “adoption metrics” like tokens per user versus prompts per week, and weigh what better reflects ongoing, valuable use.
• Justin describes where the governance battles are emerging now: permissions, agent access patterns, AI clauses in contracts, and how to build an internal org chart that can manage AI agents like a new

💡 Key Takeaways

• “AI is real” is not the same thing as “AI is widely used.” Real adoption shows up when an organization can safely incorporate it into workflows and data boundaries.
• The pause is partly rational: once AI is embedded, the limiting factor becomes governance, not novelty or access.
• Token usage is a tempting metric, but it can reward inefficiency and does not necessarily correlate with value, especially in consumer scenarios.
• The biggest operational bottleneck is org-wide alignment: you can token-max development, but ROI still collapses if the rest of the company cannot keep up.

❓ Questions That Mattered

• How do we differentiate early adoption by curious consumers from sustained, workplace-relevant adoption inside organizations?
• Which measurement is most honest: tokens per user, prompts per week, time-in-platform, or something else that reflects real value over time?
• What does “success” even mean after the novelty phase, when policy, governance, security, and data access are now the gating factors?
• Are there governance solutions that can unlock cross-silo collaboration without creating new unacceptable risk?

🗣️ Notable Quotes

• “99 % of those 1.3 billion individuals that are using AI currently are just using AI through a free feature, a free account.”
• “it’s more than just buy a license and tell people to use it.”
• “we’re kind of in this pause moment where organizations, leaders, boards, managers, directors, employees are all identifying, holy cow, okay, the tool's really powerful.”
• “You’re as fast as your slowest team.”

🔗 Links & Resources

• Listen To Other Episodes Co-Hosted With Justin