Adam Grant on How AI Reshapes Work: Why Agility Beats Ability

Why It Matters

Because AI changes the speed of knowledge creation, companies that prioritize adaptability and systematic experimentation will sustain competitive advantage, while those clinging to legacy ability‑based metrics risk obsolescence.

Key Takeaways

• Agility, not ability, is the new success metric in AI‑enabled teams
• Leaders should run frequent low‑risk AI experiments and compare human‑AI workflows
• Rewarding failure and sharing AI‑derived insights boosts psychological safety
• Structured “challenge networks” using AI help expose strategic blind spots
• Pair autonomy with innovation tournaments and kill‑signals to avoid AI chaos

Pulse Analysis

In the post‑pandemic era, artificial intelligence has accelerated the pace at which information is generated and decisions are made. Grant’s assertion that “agility is the new currency of success” reflects a broader shift from static expertise to dynamic learning ability. Companies that can rapidly test prompts, compare human‑AI outputs, and unlearn outdated practices gain a decisive edge, while firms that rely on entrenched skill sets find themselves lagging behind. This mindset aligns with recent Gartner forecasts that 70 % of AI projects will fail without a culture of continuous experimentation, underscoring the strategic imperative to embed adaptability into every workflow.

Grant proposes concrete mechanisms to operationalize agility. Running low‑risk, hypothesis‑driven AI experiments mirrors the scientific method and creates a feedback loop that refines both prompts and processes. Structured “challenge networks,” where AI is asked to surface blind spots, complement traditional 360 reviews and surface strategic gaps faster than human‑only panels. Innovation tournaments such as Atlassian’s ShipIt or Dow’s prize‑funded idea contests provide a sandbox for cross‑functional ideas, while predefined kill‑signals—borrowed from Google X—ensure resources are withdrawn early from dead‑end projects. Together these practices turn AI from a novelty into a disciplined growth engine.

The cultural layer is equally critical. Stories that celebrate junior employees who pilot AI use cases and senior leaders who openly share AI‑identified weaknesses reinforce a norm of psychological safety and learning from failure. Reward systems must shift from volume‑based output to concise, insight‑driven contributions, mitigating the “work slop” that AI can generate at scale. By redesigning rituals, performance metrics, and information flows—much like WL Gore’s lattice structure—organizations prevent the ossification of legacy practices and keep the workforce responsive to rapid AI evolution. Firms that master this holistic approach will convert AI’s speed into sustainable, high‑impact performance.