What skills survive AI disruption? Raoul Pal, macro investor and founder of Real Vision, has an answer, and it starts with a provocation most professionals are not ready to hear: knowledge is now worth zero.
Not diminished. Not undervalued. Zero.
He was not being provocative. He was being precise. When any question can be answered in seconds, when any body of information can be summarized, analyzed, and cross-referenced by a system that never sleeps, the competitive value of knowing things, as a standalone capability, has effectively collapsed.
That is not a prediction about where we are headed. It is a description of where we already are.
And it changes everything about what professionals should be building right now.
What Got Disrupted
For most of professional history, knowledge was the moat. You studied for years to build it. You practiced to deepen it. The person who knew more, who had read more, processed more, retained more, held a compounding advantage over the person who had not.
Law. Medicine. Finance. Marketing. Engineering. Accounting. Every professional field was organized around the accumulation and application of specialized knowledge. That is what credentials certified. That is what experience confirmed. That is what clients paid for.
AI did not simply accelerate access to knowledge. It commoditized it. The gap between someone who has spent twenty years accumulating expertise and someone with access to a well-trained AI model is now, for a wide range of knowledge-retrieval tasks, negligible.
That is the disruption. It is not coming. It arrived.
What the Data Shows
Researchers at MIT ran a controlled study of professional scientists. The ones using AI tools became 44% more productive on routine tasks. But here is what the study also found: the productivity gain was almost entirely in knowledge retrieval, synthesis, and documentation, tasks that follow a known path.
The scientists who performed work that required original hypothesis formation, unexpected connection-making, and judgment about which questions were even worth asking? AI did not close that gap. In several cases, the AI-assisted scientists performed worse on those tasks, because the tool had optimized their time toward execution and away from thinking.
Knowledge retrieval got cheaper and faster. Original thought became more valuable, not less.
That is not a coincidence. It is the pattern.
What Skills Survive AI Disruption: The Five That Are Not Worth Zero
The 7-Sided Pincer Movement is the set of forces compressing professional value right now. The 5-Ingredient Kryptonite Defense is the answer to it. And the reason the defense works is exactly what Pal’s observation reveals: the five ingredients are the capabilities that AI cannot retrieve, synthesize, or deploy on your behalf.
These are not soft skills. They are not old-fashioned virtues. They are the specific human capabilities that remain above the line AI has drawn, and they are the answer to what skills survive AI disruption at the individual, team, and organizational level.
IDEAS. Not information. Not the ability to recall what others have thought. The ability to generate original insight, to see a connection no one has named, to ask the question the room has not asked, to propose the frame that makes the problem solvable. AI can recombine. It cannot originate.
SPEED. Not processing speed, AI wins that race permanently. Decisional speed: the ability to act on incomplete information, to commit under uncertainty, to move before the full picture is available. Organizations that build decisional infrastructure, the systems, culture, and leadership that allow fast judgment calls, cannot be replaced by faster hardware.
TALENT. The human capabilities above the line that autonomous systems draw. When Helena, an autonomous AI marketer, can build a full brand strategy, run Meta ads, and publish SEO content overnight for any business, the marketing coordinator below that line is exposed. The strategist above it, the one who knows what story to tell and why it matters to this specific customer, is not.
DISTINCTION. The ability to be unmistakably, irreducibly yourself, as a professional, as a brand, as an organization. AI produces competent averages. Distinction is not average. The professional whose perspective, voice, and judgment are genuinely their own occupies territory no model can colonize.
LEADERSHIP AT ALL LEVELS. The ability to move other humans, to build trust, to navigate conflict, to hold teams together under pressure, to develop the people around you. AI can report on team performance. It cannot earn the room.
These five capabilities are not AI-resistant because they are old or sentimental. They are AI-resistant because they require something AI does not have: a self. A stake. A relationship. A history. The judgment that only comes from having something to lose.
What This Means for You Right Now
The professionals most at risk in the current disruption are not the ones who know the least. They are often the ones who built their entire competitive identity around knowing the most and have not yet rebuilt it around something AI cannot replicate.
The question is not whether your field is being disrupted. Every field is. The question is which parts of your work live above the line that AI has drawn, and whether you are actively building more of your work in that territory.
Understanding what skills survive AI disruption is not an academic exercise. It is the most important career and organizational strategy decision of this decade.
Knowledge was always the floor. It was never supposed to be the ceiling.
If AI has collapsed the floor, the ceiling is now all that matters.
Those prepared need not fear the forces at work.
Want to find out which of the five ingredients is your strongest, and where you and your organization are most exposed? Take the Kryptonite Scorecard at realmikeevans.com/scorecard.
Related: The AI Layoff Trap: Nobody Is Doing Anything Wrong.
Related: AI Made Some Scientists 44% More Productive. And Did Almost Nothing.