The High Agency Era: Why AI Rewards the Self-Directed

The career ladder is being disassembled while people are still standing on it. Entry-level hiring at major tech companies has collapsed by more than 50% since 2019. The routine tasks that once trained newcomers are precisely the tasks that generative AI handles with increasing competence. This is not a temporary hiring freeze. It is a fundamental restructuring of how careers work.

The rungs are gone. The climb just got steeper. And the trait that now separates those who will thrive from those who will struggle has a name: high agency.

The Trait That Matters Now

Psychologist Julian Rotter identified the concept in the 1950s: locus of control. People with an internal locus believe they primarily control their outcomes. People with an external locus believe outcomes happen to them.

Here is the exercise that reveals where you stand. Draw a circle on a piece of paper. Write down everything that matters: career, family, learning, promotion, goals. Whatever falls inside the circle is what you perceive as under your control. Whatever falls outside is beyond your influence.

When high agency people hear that voice suggesting something is beyond their control, they respond with four words: That's a skill issue. Not a permanent limitation. A skill issue. Something they can learn. Something they can figure out.

The Research Is Clear

The data on internal locus of control is remarkably consistent:

• Greater persistence when facing challenges
• Higher academic achievement
• Better professional outcomes
• More effective leadership
• Stronger entrepreneurial activity

Students with an internal locus outperform those who attribute outcomes to external forces by 20 to 30% on average. Companies led by CEOs with an internal locus are more likely to survive difficult periods. Angela Duckworth's research on grit found it to be a better predictor of success than IQ in contexts ranging from West Point cadets to spelling bee contestants.

This is not about delusion. High agency people do not pretend obstacles don't exist or that the world is fair. They simply recognize there is an uncertain number of obstacles between them and any significant goal, and they maintain an irrationally confident belief those obstacles can be overcome.

AI as the Great Accelerant

Here is where it gets interesting.

Before this moment, whether you had high agency or not sorted itself out over the course of decades. The signal was gradual. You could coast for a bit. You could wait for the next rung to appear.

That world is gone.

Andrej Karpathy, former chief scientist at OpenAI, recently admitted he "had this intuitively wrong for decades" due to our cultural veneration of IQ. His conclusion: "Agency is significantly more powerful and significantly more scarce than intelligence."

Google engineering lead Addy Osmani frames it this way: "In a world full of brilliant ideas and talented people, the real scarcity is those with the agency to execute and make ideas happen."

Graham's thought experiment: take a person who is 100/100 in both intelligence and determination. Dial down their determination and you get a genius who can't get anything done. Dial down their intelligence while keeping determination maxed, and you get a moderately smart business owner who still manages to get rich.

The Intelligence Trap

If intelligence is so great, why do so many smart people fall short?

Analysis paralysis. The more brainpower you have, the more adept you become at imagining everything that could go wrong. Bright minds overanalyze, seeking optimal solutions or perfect certainty. The result: they never act.

Carol Dweck's research found that people praised solely for being "smart" often develop an aversion to challenges. They don't want to risk looking dumb or losing their smart status. The child who's always been the smartest in the room crumbles at a task that doesn't come easily, because they never learned to push through struggle.

Meanwhile, the determined "B student" leaves the genius "A student" in the dust. Not because intelligence doesn't matter. Because intelligence without action is like a car with no fuel.

The Say-Do Ratio

The second component of high agency is what might be called the say-do ratio.

If you say you're going to do something, do it. Not tomorrow. Not after making a plan. Not after talking about it. Now.

Most people have a terrible say-do ratio:

• They say they'll learn a new skill, then research courses for three weeks
• They say they'll start a side project, then spend a month choosing the perfect tech stack
• They say they'll get in shape, then spend evenings reading about optimal workout routines

The gap between intention and action stretches into weeks and months.

There is extraordinary progress available when you collapse the distance between saying you'll do something and actually doing it. This often means starting when it feels uncomfortable. Beginning before you feel ready. Shipping halfway done rather than waiting for perfection.

AI Changes the Equation

Here is the piece that changes everything.

AI is the greatest equalizer for agency that has ever existed.

The barriers that used to require years of expensive education or access to just the right networks can now be overcome by someone with a laptop and the determination to figure it out. AI doesn't care about your pedigree. It responds to your questions.

Before AI, if you wanted to be a solo founder but couldn't code, it was a massive obstacle. Years of learning. Now you move through that obstacle in weeks or days. You can build functional prototypes while learning. The same applies to marketing, content creation, data analysis.

The Recruiting Signal

Dr. John Sullivan, who Fast Company called "the Michael Jordan of Hiring," now argues that companies should specifically recruit for high agency capabilities. His reasoning: those with high agency solve complete problems rather than narrow slices. They target high-cost strategic problems. They continuously learn. They come with proven track records.

More importantly: they don't require close management. They're purpose-driven and willing to take complete ownership.

The characteristics he tells recruiters to look for:

• Evidence they've influenced their environment rather than accepting it as given
• Optimistic problem-solving approach
• Self-directed continuous learning
• Data-driven decision making
• Developed disruptive solutions rather than accepting best practices
• Tenaciously attacked problems without fear
• A sense of urgency

Sound familiar? It's the same profile that succeeds with AI tools.

The Mindset Shift

What's interesting (and rarely discussed) is that if you approach life with high agency, your approach to AI automatically becomes dramatically better.

The breathless articles about students using AI as an answer engine, about losing critical thinking, about IQ points declining from AI exposure: all those outcomes happen when people are passive. They happen when people lack agency.

If you are high agency, you engage with AI differently. You use it to explore problems more deeply than you could alone. You use it to learn skills faster. You use it to prototype ideas.

What This Means

The career ladder as it existed is gone. But something potentially more powerful has emerged in its place: a direct link between individual agency, intelligent tools, and the ability to create value at scale.

The people who see this clearly are already moving. They're building. They're shipping. They're learning. They're iterating. They're not waiting for permission or for the old structures to return.

Because they're not coming back.

Sources: Nate B Jones (AI News & Strategy Daily), Addy Osmani, Dr. John Sullivan, McKinsey "Superagency" research, Julian Rotter (locus of control), Carol Dweck (mindset research), Angela Duckworth (grit), Paul Graham, Derek Sivers.