When Code Is No Longer Leverage

Naval once said code and media are "permissionless leverage." This statement inspired countless indie developers and content creators over the past decade.

But here's the problem: if a single prompt can build a product, if AI can write 100 articles in 10 seconds, are these still leverage?

The answer: when everyone has leverage, leverage itself no longer creates relative advantage. It degrades from "weapon" to "survival necessity."

This isn't denying AI's value. It's saying that if you're still thinking like it's 2023—writing articles and building products the old way—you'll drown in the digital flood of AI-generated content.

Your Hustle Is Being Diluted by AI

Traditional thinking goes: AI increases productivity, so I should use AI to produce more.

Reverse thinking goes: since AI explodes the supply of digital products, "production" itself is depreciating.

This doesn't mean don't use AI. It means what you use AI for matters more than whether you use it.

A brutal reality: the article you wrote with AI might, at the exact second you publish it, compete with 1,000 similar pieces being batch-generated across the web. Your "diligence" is being diluted by AI's "diligence."

Code and Content Are Just Entry Tickets Now

Naval said humans have four types of leverage: labor, capital, code, and media.

In the AI era, "code" and "media" are degrading from leverage to infrastructure. They're no longer advantages—they're entry tickets.

Of the remaining two, "labor" has too high a barrier for individuals, and "capital" is out of reach for most.

So what's the new leverage in the AI era?

The answer: Accountability (responsibility/reputation) and Specific Knowledge (proprietary knowledge).

AI Can't Go to Jail—That's Your Opportunity

AI won't go to jail. AI doesn't bear legal responsibility. AI doesn't suffer from failure.

This means in healthcare, law, core financial decisions—fields requiring "full accountability"—human guarantee becomes the most expensive leverage.

More broadly, when the market floods with AI-generated content and products, a real human willing to stake their reputation (skin in the game) captures the maximum trust premium.

You might say: I'm not a doctor or lawyer, what's this got to do with me?

Here's the connection: even if you're just blogging or building small tools, "I personally verified this" versus "AI generated this for me" hold completely different values in users' minds.

Do What AI Can't Google

AI's data source is the public internet. Information not on the internet is forever inaccessible to AI.

This includes: industry unspoken rules you've touched firsthand, your private relationships with key players, real pitfalls you've encountered in niche verticals.

None of this exists in any public dataset. AI cannot imitate it, cannot replicate it.

Reverse move: stop using AI to produce "general knowledge" content. Dig deep into a niche vertical AI hasn't polluted yet (preferably with offline elements). Use AI to 10x that industry's efficiency instead of creating another AI tool on the internet.

Stop Being a Carpenter, Be a Broker

Many developers have a fatal habit: obsessing over making products more perfect, adding more features.

This is technical thinking. Technical thinking sees yourself as a carpenter, focused on making the chair prettier.

But in the AI era, chairs can be mass-generated. Carpentry skills are no longer scarce.

Business thinking is completely different. Business thinking doesn't care who makes the chair. Business thinking's core is: discover demand, intercept traffic, match transactions.

A painful conclusion: The product is just a giveaway for traffic.

If a product can't show users "pain solved" within 10 seconds, it's commercially worthless scrap metal.

Kill Mediocrity to Gain Compound Returns

A founder posted this on Twitter:

"Founders keep telling me AI made them busier. 'I can build so much now!' Then: 'I'm exhausted.' Of course. They used AI to add. Could've used it to subtract. I killed more than I built last month. Most peaceful month I've had."

This hits the biggest trap of the AI era: the addition trap.

AI lets you write 10 articles a day, build an app in an hour. You enter "productivity euphoria," feeling like you're progressing. In reality, you're just accelerating the production of mediocrity.

Reverse move: use AI for subtraction, not addition.

Kill features without conversion. Kill soulless content. Kill AI-flavored bullshit.

A blog with 10 "masterpieces" outweighs a site with 1,000 "AI slop."

"What you don't publish" shows your level more than "what you publish."

Don't Teach People How—Help Them Avoid Pitfalls

Traditional "How-to" content has become cheap. Users no longer need you to teach them how (AI can teach)—users need certainty.

Reverse move: don't be a "generator," be an "auditor."

AI handles "producing junk" (generating content/code/resumes/plans). You handle "judging quality" (diagnosing risks/avoiding pitfalls/auditing).

"Helping others avoid pitfalls" commands higher prices than "giving others tools." Because the former involves fear, the latter only involves efficiency.

An interesting observation: in the AI era, "taste" becomes a scarce resource. AI has no taste—it only has averages. When AI can write 100 articles scoring 80 points each, users need someone to tell them "which one is worth 5 minutes of reading."

You're earning money for "saving users' mental bandwidth," not for "providing information."

Stop Seeding, Start Fishing

Many treat articles and apps as "seeds," believing if they plant enough, something will sprout.

This is seeding thinking. In the AI era, seeding thinking is going bankrupt. Because the soil is already covered with AI-generated weeds.

Reverse move: upgrade from seeding to fishing thinking.

Every article, every landing page must be a "hook." Articles without hooks are wasted electricity.

A sharp headline + a truth-telling opinion + a Contact Form that solves pain = fishing gear.

More importantly: intent to transact first, product second.

If your hook gets no bites, no matter how perfect you make the product, you're just fishing for air in the open sea.

Three Reverse Paths

Since "production" is depreciating, what's appreciating?

Path One: From "owning code" to "owning proprietary data and physical loops"

If a single prompt can build an app, that app's logic is transparent and easily copied. Reverse logic: since software is no longer a moat, "heavy assets" or "offline" become the moat.

AI can write you the perfect warehouse management system, but it can't secure your location, your fleet, or your difficult suppliers. When the digital world goes extremely "light," whoever can root AI's output in dirty work, hard work, legal barriers, and physical moats owns the premium.

Path Two: From "creating content" to "curation and trust endorsement"

Since AI can write 100 articles scoring 80 points, users need "tell me which one is worth my 5 minutes." Future influencers won't be content producers—they'll be "filters."

You earn money for "saving users' time," not for "providing information."

Path Three: From "building products" to "system architecture and complex orchestration"

"One prompt = one product" usually makes a simple island tool. Real-world problems are complex and long-chain.

Standalone tools lose value. The ability to chain 100 AI tools together to solve complex business problems gains value. Example: don't build an AI writing tool—build a complete automation matrix from "trend monitoring → auto-rewriting → multi-platform distribution → comment auto-reply → conversion analytics."

80% of Your Time Should Be Spent Selling

Stop using AI to produce "general knowledge" content or products.

Dig deep into a niche vertical AI hasn't polluted yet (preferably with offline elements). Use AI to 10x that industry's efficiency instead of creating another AI tool on the internet.

More specifically:

Spend 20% of your daily time on "building with AI" (don't aim for perfect, just functional).

Spend 80% on "writing sharp traffic-intercept articles, manually intercepting in social media comments, optimizing your sales copy."

This is reverse thinking: treat products as "giveaways," treat traffic and closing as "core business."

Slow, Heavy, Painful, Clumsy—That's Your Moat

AI increased our productivity but stole our compound returns.

The only way to reclaim them is to return to that real flesh—the one that hurts, makes mistakes, and dares to take responsibility.

"Slow," "heavy," "painful," "clumsy"—these traits once seen as inefficient have become the highest-level moats in the AI era.

When you feel AI-written articles are cheap, you should hand-write those words carrying "pain" and "sweat."

That's the true essence of reverse thinking.

Further Reading

AI Marketing Automation

How to Use Vibe Marketing for AI-Automated Marketing

How AI Can Do Your Work Automatically

This One Tool Can Revolutionize 80% of Your Workflow

AI Cognition

The End of Copying

Maxwell's Demon Guide

The E^F Law

AI Monetization

Can't Make Money with AI Domestically—Must Go Global?

AI Monetization: Building Demo Sites for Bosses

Tutorial Series

Prompt to Product: Earn USD Going Global