Don't Cognitively Surrender Your Human Advantage
Ezra Klein’s recent essay in the New York Times (subscription required) outlines a phenomenon people should pay attention to. In his recent conversations with AI leaders, Klein says he picked up on the consistent theme that these individuals are deeply integrating AI tools into their lives, essentially handing over more and more cognitive processes to Claude, Codex, Gemini et al.
Within this trend, Klein draws a distinction between “cognitive offloading” — handing discrete tasks to AI so you can focus on more important work — and “cognitive surrender”, which is what happens when you stop thinking deeply and instead “surrender” the hard part of work to your AI companion as well. Klein’s punchline is simple enough. Struggling through a hard problem is exactly what deepens human thinking and, in turn, builds expertise that can be applied when needed in the future.
Though Klein maps the concept of AI-fostered “cognitive surrender” to his own journey as a writer, it can easily be applied to the workplace where it might impact your role as a leader. When your team members use AI to skip the struggle, the near term output can be done quickly and, at first glance, looks good enough. A strategy deck gets produced, a financial model gets built, and a marketing plan lands in your inbox. But long-term you get the same results as when your body scarfs down a bunch of empty calories. Yep, you get a short-term high that comes from getting the work done so quickly, but what isn’t happening is the development of real expertise and judgement. Your team members are surrendering those important benefits to the LLM not just today, but also in the future.
And if you poke at those AI-generated results a few times you’ll fully recognize they’ve been crafted without the durable foundation needed to benefit your organization long-term. Examples you might come across include:
the customer acquisition strategy presentation AI created in minutes, but assumptions behind the plan can’t be defended by the marketing director
the product roadmap plan that AI put together but the customer problems being prioritized to solve can’t be clearly articulated by the product manager
the financial model whose revenue build came right from AI but the questions asked by board members can’t be adequately responded to by the finance team leader
The output is there, but the depth of expert understanding isn’t.
So as a CEO, functional department leader, or someone working anywhere in an organization, what should you protect when it comes to owning your cognitive development? Start with these five domains where doing the hard work will yield durable human expertise versus surrendering it to AI:
Setting goals and defining OKRs that define how to drive your organization forward
Outlining and fully understanding the core elements of your product, customer acquisition, and overall business strategy
Learning how to recognize when a strategy is being executed well, and when it isn’t, so course corrections and “reforecasts” can be developed in real time
Developing partnerships with other humans and learning how to negotiate those relationships to support the operating goals for your company
Calculating return on both people and technology investments, and being able to defend those calculations to your executive team and board of directors
These are not tasks to delegate to an AI agent. They require a deep understanding of how your business works and the strategy you’re pursuing, and how to apply that understanding in ways that enable your organization to be successful. Truly developing the skills above (and others like them) defines whether your team is actually learning and can continue to adapt, or whether they’re just learning to prompt this competitive advantage away in conversations with AI.
Starting this week, pick one meeting where your team presents strategic work and ask the presenters to walk you through their thinking, not just their output. Push on the assumptions. See who can go deeper with you as you ask more specific questions. Use this session to shine a light on examples where some “cognitive surrender” may be creeping into your organization, and to also highlight examples where “doing the hard work of thinking” shows up so your team can clearly see the important differences between the two.
NextPlay>Forward AI Disclaimer: I very actively use artificial intelligence and large language models to generate the content you read here, but I do review it and edit it to make sure it can be generally useful to people who read it. Keep in mind that AI can make mistakes - check important information. Let me know if I make any errors and I will correct them.