AI Needs Us More Than We Need It

95% of organizations using AI are getting zero return on their investment.

Let that sink in.

MIT studied this.
Goldman Sachs confirmed it.
Atlassian’s research backed it up.
Billions spent on AI in 2025, and almost nobody’s actually making money with it.

You know what that tells me?

The “Average” Trap

A study dropped in January 2026. Scientists tested 100,000 humans against the best AI on creativity tasks.

The results were revealing.

AI beats the average person now. If you’re mediocre, AI can replace you. That’s the hard truth.

But here’s what didn’t make the headlines: the top 10% of humans still crush every AI system tested.

The top half of humans scored higher than every single AI.

So AI isn’t replacing creativity. It’s filtering for it.

What 90% of AI Content Actually Is

YouTube’s CEO called it out in January: “AI slop.”

That’s his term for the flood of garbage content hitting the internet. Cheap. Mass-produced. Empty calories for your brain.

By the end of 2026, 90% of web content could be AI-generated.

Think about that. Nine out of ten things you read online might be synthetic. Generated without thought. Published without care. Content for content’s sake.

And you know what’s happening simultaneously?

Platforms are desperate for it. Users are starving for it. In a world drowning in AI slop, the person who can find the signal becomes royalty.

The work didn’t disappear. It transformed.

We went from content creators to content curators. From producers to quality controllers. From makers to judges.

And judging—real judgment, taste, discernment—that’s a deeply human skill.

The Hallucination Problem Isn’t Getting Better

I know what the AI companies promised you. “Better models.” “Reduced hallucinations.” “More reliable outputs.”

Here’s the reality: hallucination rates still run 5% to 40% depending on the domain.

GPT-4o—their flagship model—showed a 53% hallucination rate on certain tasks. You could improve it with better prompting. Get it down to 23%.

But that’s still nearly one in four outputs being wrong.

You want to hand your business decisions to a system that’s wrong 25% of the time?

Didn’t think so.

Every AI output requires human verification. Every “fact” needs checking. Every recommendation needs context that AI doesn’t have.

The work isn’t gone. It’s just moved upstream.

Instead of writing the report, you’re fact-checking it. Instead of building the model, you’re validating it. Instead of doing the work once, you’re doing it twice: once by AI, once by human.

And that second pass—the human judgment layer—that’s where the value lives now.

My Real Experience With AI Leverage

I want to tell you what’s actually happening in my business.

I’ve got AI tools everywhere. Writing assistants. Image generators. Code copilots. Analysis tools. I’ve tried them all.

And you know what? I’m working harder than ever.

But it’s different work.

I used to spend hours on first drafts. Now AI generates them in seconds. But then I spend more time refining, redirecting, adding the human element that makes it actually good.

And here’s the weird part: it feels better.

Way better.

Because I’m not doing the busywork anymore. I’m not formatting documents. I’m not writing routine emails. I’m not grinding through administrative tasks that drained my soul.

I’m doing the creative work. The strategic work. The work that requires me to be human.

The AI Twins Effect

James Caan (the recruiter, not the actor) published something interesting in January.

He analyzed “AI Twins”—professionals using AI avatars of themselves to handle routine work.

Recruiters using this tech save 12 to 15 hours per week.

That’s nearly half a work week.

Here’s what the successful ones do with that time: they don’t do more volume.

They do better work.

They build deeper client relationships. They think strategically. They create trust.

That’s the whole game right there.

It’s not human vs. machine. It’s augmented human vs. unaugmented human.

The winners aren’t the ones using the most AI. They’re the ones using AI to become more human, not less.

The Real Work Is Process Design

Research from 2026 shows something counterintuitive.

Teams using AI assistants gain 10-15% productivity. But only if they redesign their workflows around the technology.

Without that redesign? The time savings evaporate. People use AI for busywork but don’t reinvest the time strategically.

High-performing teams—the ones seeing 16-30% productivity gains—do something different. They completely rethink how work happens.

They ask: “If AI can do X, what should humans do instead?”

And the answer is always: higher judgment work.

Strategic thinking. Creative direction. Relationship building. Quality control. Taste-making.

The new work isn’t execution. It’s orchestration.

You don’t write the symphony anymore. You conduct it.

And conducting—knowing when to bring in the strings, when to quiet the brass, how to shape the emotional arc—that requires a human who’s developed taste and judgment over decades.

The Creativity Paradox

Here’s the deepest truth about AI and creativity:

AI creativity depends entirely on human input.

That same study with 100,000 participants? They tested how AI responds to different prompting styles.

When you encourage AI to think about etymology and word origins, it generates more unexpected associations and higher creativity scores.

The creativity isn’t in the machine. It’s in the interaction between human and machine.

The prompt engineer matters more than the model. The curator matters more than the generator. The judge matters more than the producer.

We’ve spent decades worried that AI would replace human creativity.

Instead, AI exposed how little of our work was actually creative.

The rote stuff? AI can do that. The average stuff? AI’s got it covered.

But the truly creative work—the synthesis, the taste, the vision, the unexpected connections?

That still requires a human.

And now that human has leverage.

What This Means For You

If you’re reading this and thinking “AI is going to replace me,” you’re asking the wrong question.

The question isn’t: “Can AI do my job?”

The question is: “What part of my job is actually worth doing?”

Because AI can probably do 80% of what you do. The routine parts. The repetitive parts. The parts that don’t require judgment or taste or human connection.

But that remaining 20%? That’s where all the value is.

And here’s the beautiful part: when you leverage AI for the 80%, you can do 5x the 20%.

More creative work. More strategic work. More human work.

Not less work. Better work.

The Renaissance Is Here

I believe we’re entering a renaissance of human work.

Not because AI is failing. Because AI is forcing us to level up.

The bar for “good enough” just got raised to “exceptional or nothing.”

Average content gets buried in AI slop. Average analysis gets outperformed by a chatbot. Average creativity gets automated into oblivion.

But exceptional work? Exceptional work has never been more valuable.

Because in a world where anyone can generate infinite content, the person with taste becomes king.

In a world where AI hallucinates 25% of the time, the person who can verify truth becomes essential.

In a world drowning in synthetic noise, the human signal becomes priceless.

This isn’t the end of human work. It’s the beginning of meaningful human work.

The busywork gets automated. The creative work gets amplified.

And yeah, it’s more work. It’s harder work. It requires more of you, not less.

But it’s the work you actually want to do.

Sources & Further Reading

• Divergent creativity in humans and large language models – Scientific Reports (January 2026)
• 95% of enterprises see zero ROI from AI – MIT Report
• Atlassian AI Collaboration Report 2025
• YouTube CEO Neal Mohan on AI Slop (January 2026)
• James Caan on AI Twins in Recruiting (January 2026)
• AI Hallucination Rates 2026
• AI-Generated Content Statistics – Graphite