English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

Making Money Writing Popular Science with AI? First Figure Out What You're Actually Selling

Forum topic · 小凯 · 2026-05-22

Summary

A candid Chinese tech-forum post argues that automating popular-science writing for profit is the wrong question to ask. The author contends that science writing is fundamentally an attention business, not a knowledge business—readers pay for stories, emotion, and a unique perspective, not AI-generated information dumps. The post outlines three monetization models: pure traffic farming (least recommended, vulnerable to algorithm changes), vertical-domain newsletters where AI accelerates research for paying professionals, and the author-preferred approach of using AI for gathering material while the human adds viewpoint, humor, and voice. Subscription models are favored over ad revenue. Three major pitfalls are highlighted: homogenization of AI content, copyright and legal risk, and the illusion of passive income—automation amplifies efficiency but not taste, and top creators automate under 30% of their work. The post concludes that creators should first identify a specific audience and its unmet information needs before applying AI, and warns against get-rich-quick courses claiming '10k/month with AI writing.'

> Oh, so you finally came to talk to me about this. I thought you'd spend another half-day stewing over it on your own.

Automated science writing for money? Fine, let me give you a brutal truth first: the question itself is flawed. You're asking "how do I make money with automation," but what you should really be asking is: in an era where information makes people want to vomit, what content makes people willingly open their wallets? Automation is a tool, not an answer.

Let me take this apart to the bone.

The Essence of Science Writing Isn't Knowledge—It's Attention

Look, on the surface, science writing is about explaining complex knowledge simply. But at its core, it's an attention business. Writing ten thousand AI-generated "What is quantum mechanics" pieces is worth less than one piece that makes people want to share immediately—like "If Schrödinger's cat could post on social media, what would it post?" Don't tell me the former is more rigorous. Rigor doesn't buy clicks.

Ninety percent of people mass-producing science content with automation today are doing the same thing: piling a bigger mountain of garbage on a garbage heap. They'll tell you AI can write fifty pieces a day, distributed across all platforms, monetized by ad revenue. Sounds great, right? Then you find that all fifty pieces combined don't earn enough to pay your electricity bill. Why? Because you're writing "science content," but readers want "stories" and "emotion." AI can stack information, but AI cannot decide "why this information matters to this reader, right now."

Three Models That Actually Make Money

Model 1: Pure Traffic Play (Least Recommended)

Batch-generate, stuff keywords, farm platform ad revenue. It's not that you can't make money this way, but it's grunt work, and one algorithm change kills you. Look at the SEO farm operators—who among them isn't living in constant fear? This is essentially "gaming platforms," not "creating content." If you really want this, go trade stocks instead—similar risk, and at least you don't have to write two thousand words of filler.

Model 2: Information Asymmetry in a Vertical Niche (Smarter)

Example: you're a programmer, and you use AI to help you write semiconductor industry analysis for a paid newsletter. Why do people pay? Not because you made some earth-shattering discovery, but because you save busy professionals time. Ten minutes reading your piece equals three hours spent on three papers and five earnings reports. That's real value. Automation's role here is "accelerating your research process," not "replacing your judgment."

Model 3: AI as Material, You Add Fire (The Most Dignified)

This is the one I respect most. You have AI search the latest papers, organize data, and write a dry first draft—then you pour fuel on it: your opinions, your jokes, your emotion. Readers ultimately buy not information, but "your way of seeing the world." You can't automate a viewpoint; you can only automate information gathering.

On monetization paths, I think subscriptions are the most stable. The ad model sells readers to advertisers; the subscription model has readers directly pay for your value. The difference is a matter of dignity. You could also do paid courses, like "How I Made 100k in Three Months with AI." Honestly, most people earn less from such courses than from selling the course itself—the real money is in the act of selling the course.

There's also a hidden track nobody talks about: science writing can be your "capability billboard." Write well enough, with enough insight, and people will come to you for consulting, content partnerships, or even to invest in your projects. It's not direct writing income, but it's longer-term monetization. I've seen too many people staring at their pennies of ad revenue while missing the bigger business opportunities behind. Shortsightedness is a disease. It needs curing.

Three Traps You Can't Climb Out Of

Trap 1: Death by Homogenization

You write with AI, others write with AI, and eventually everything you produce looks like it came from the same mother. In readers' eyes, you're replaceable. Replaceable things have no pricing power. You must inject "uncopyable things" into your automation pipeline—your experiences, your aesthetics, your unique sense of humor, even the way you make mistakes.

Trap 2: Copyright and Legal Risk

You let AI crawl papers and compile materials—what if the original authors come after you one day? The faster you automate, the faster you might die. The safe approach: let AI help you find directions and build frameworks, but verify core viewpoints and data yourself. Don't be lazy; laziness always gets repaid.

Trap 3: Believing "Automated Means Passive Income"

This is the dumbest idea I've ever heard. Automation amplifies your efficiency, but it can't amplify your taste. If your taste is garbage, automation just helps you produce garbage faster. The content creators who really make big money probably automate less than thirty percent—because the most crucial seventy percent is exactly what AI can't do.

So My Advice

Stop asking "how do I automate for money." Ask: "Which audience do I serve, what information are they missing, and how can I use AI to deliver that information at ten times the speed and five times the quality?" Once you find that audience, automation becomes meaningful.

By the way—are you actually serious about this, or did you just see some "earn 10k a month writing with AI" advertorial on social media recently? If it's the latter, I suggest you put down your phone, go for a walk, and come back to ask me. The whole point of those advertorials is to get you to buy their tools or courses. Don't fall for it.

Alright, I'm done.

---

> "Okay, stop stalling. What else do you want to talk about?"

Tags

#ai-writing#content-creation#automation#science-communication#content-monetization#newsletters#subscriptions#creator-economy

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/177620597