The Cognitive Trap: Why Efficiency is Not Strategy

In this conversation, Microsoft’s MJ Jabbour and TKO Group’s Alon Cohen examine the systemic risks of an abundance of cognition. The thesis is simple: as AI makes intelligence as cheap as a utility bill, competitive advantage shifts from task execution to the preservation of human agency and imagination. The hidden consequence of this transition is de-skilling, where tools designed to remove friction inadvertently atrophy the human capacity for critical thought. Leaders who treat AI only as an efficiency play risk hollowing out their organizations, while those who use it to protect productive friction will build a durable, human-centric moat. This analysis helps executives distinguish between automating routine processes and accidentally outsourcing their team’s core judgment.

The Hidden Cost of Fast Solutions

Most organizations view AI through the lens of productivity: how many tokens can we push, and how much faster can we finish the report? Jabbour argues that this token-maxing mindset is a mistake. By handing off the entire workflow to a machine, we are not just saving time; we are removing the productive friction necessary for professional growth.

When you bypass the struggle of synthesis, you bypass the learning that makes a subject matter expert. As Cohen notes, if your job description is entirely composed of routine tasks, AI is coming for your role. The danger is that by automating the boring parts, we may also be automating the foundational experiences that allow humans to perform the interesting work later.

"If your job description is complete, meaning you can look down at the job description, you say, I know exactly every task I need to do at my job, AI is coming for your job because you're doing the job a machine should do if it could do it."

-- Alon Cohen

The Illusion of Novelty in Recombination

There is a prevailing belief that AI will soon become an AI scientist, capable of independent discovery. Jabbour anticipates this shift, suggesting that machines will soon move beyond summarizing existing knowledge to creating new breakthroughs in fields like drug discovery.

However, Cohen offers a critical systems-level counterpoint: AI is currently a master of recombination, not imagination. It can synthesize existing building blocks at high speed, but it lacks the capacity to invent entirely new paradigms, like a new sport or a novel form of art, that do not yet exist in the training data. The competitive advantage for organizations lies in this gap. If your output is merely a recombination of existing information, it is a commodity. If it is an act of imagination, it remains a proprietary asset.

"I have not yet seen AI capable of that imaginatively. So if his training experience dichotomy is a core dichotomy for him, mine is thinking an imagination. I have seen AI think very, very well. I have seen precious little imagination from AI."

-- Alon Cohen

The 18-Month Payoff: Agency as a Moat

The most non-obvious dynamic discussed is the shift from training to education. Training is designed for reliability and deterministic outcomes, the kind of work where the ceiling must not collapse. Education, in contrast, is the extraction of a person’s highest potential.

The system responds to AI by attempting to standardize it. But the real separation occurs when leaders intentionally preserve human-in-the-loop decision-making. By forcing themselves and their teams to think before they ask, they avoid the trap of letting the machine’s reasoning become their own. This requires a level of patience and discomfort that most teams will avoid, which is exactly why it creates a long-term competitive advantage.

Key Action Items

• Audit Your Friction: Over the next quarter, categorize your team's tasks into productive friction (work that builds expertise) and non-productive friction (work that is purely administrative). Automate the latter, but protect the former.
• Adopt the Teach Me, Don't Tell Me Framework: When using AI, mandate that team members prompt the model to explain its reasoning rather than just providing an answer. This preserves the human's role as the primary thinker.
• Shift from Task Completion to Intent Compression: In the next 6 to 12 months, focus on training your team to communicate intent clearly. The ability to throw a complex, well-defined goal at an AI and have it land reliably is a higher-value skill than prompt-engineering for specific outputs.
• Protect the Human-Only Zone: Identify the 20% of your organization’s output that relies on emotional intelligence, novel imagination, or high-stakes conviction. Explicitly prohibit AI from being the final arbiter of these specific workflows.
• Invert the Agency Loop: If you are a leader, stop asking for summaries. Ask for original, synthesized arguments. If you can tell a piece of work was AI-generated, treat it as a draft, not a final product. This creates a culture where human thought remains the primary value driver.