Fast Thinking Teams Need Autonomous Full-Stack Thinkers

The "Fast Thinking vs. Slow Thinking" Organizational Model for AI Innovation

In the AI era, companies need to restructure how they organize teams and execute on product development. Howie Liu's experience transforming Airtable offers a powerful framework for balancing rapid innovation with sustainable growth.

The Fast Thinking vs. Slow Thinking Team Structure

Fast Thinking Teams

• Focus on shipping new capabilities on a near-weekly basis
• Prioritize jaw-dropping value and user experience
• Operate with high autonomy and minimal process
• Embrace experimentation over detailed planning
• Optimize for rapid learning and iteration
• Responsible for creating top-of-funnel excitement and new use cases
• Officially called "AI Platform" at Airtable

Slow Thinking Teams

• Focus on deliberate, premeditated bets that require careful planning
• Build infrastructure and scalability features that can't be shipped quickly
• Handle complex data challenges and enterprise requirements
• Allow initial adoption seeds to grow into larger deployments
• Complement fast thinking teams by providing stability and scale

Why This Model Works

• Fast thinking creates excitement and drives adoption
• Slow thinking ensures durability and expansion over time
• The combination addresses both innovation and scale challenges
• Prevents the common AI startup problem of high initial interest but poor retention

The IC CEO: Leaders as Individual Contributors

• CEOs and leaders need to get back into the details in the AI era
• Being in the weeds is necessary to understand what's possible with AI
• Leaders should be "chief taste makers" who understand product possibilities
• This requires direct, hands-on experience with AI tools and models
• Cut standing one-on-ones in favor of timely, insight-driven meetings
• Focus on urgent topics rather than routine check-ins
• Spend time directly using AI tools and prototyping solutions
• Share examples of your own AI experimentation to lead by example

Skills for Product Teams in the AI Era

• Every role needs to become more versatile and cross-functional
• Product managers need to become hybrid PM-prototypers with design sensibilities
• Engineers need product thinking to understand user experience implications
• Designers need technical understanding to work with AI capabilities
• Success depends more on individual attitude than role definition
• The advantage goes to "polymaths" who can cross between disciplines
• Everyone should develop a minimum baseline in adjacent disciplines
• Encourage "play" with AI tools to develop intuition for possibilities

Execution Principles for AI-Native Companies

• Prioritize experimentation over detailed planning
• Build prototypes instead of writing documents
• Show, don't tell - demonstrate value experientially
• Focus on vibes before evals for novel product experiences
• Use evals once you've converged on the basic scaffold and use cases
• Collapse role silos across the entire organization
• Encourage everyone to become more "full stack" in their approach
• Be outcome-oriented rather than process-oriented

How to Become AI-Native as a Leader

• Use AI tools hourly, not just daily
• Be "intentionally wasteful" with inference costs for high-value insights
• Try as many different AI products as possible, not just your own
• Create personal side projects to force deeper engagement with tools
• Understand both the models and the product form factors
• Share your experiments and learnings with your team
• Encourage teams to cancel meetings and dedicate time to AI exploration
• Ask "how would an AI-native company execute?" and match that pace