"AI won't replace you, but people who use AI will."
That line captured the mood of early generative AI. Excitement, fear, and a lot of noise. Everyone rushed to try the latest viral prompt or tool, then quietly admitted, “Cool, I tried it,” and never touched it again.
What most people were missing was not access to tools. It was a clear signal about what actually mattered and how to use AI in real work. That gap became the opening for The Rundown AI, a daily AI newsletter startup that went from zero to over 2 million subscribers in about two years.
This is the story of how founder Rowan Cheung built that growth engine, why he bet everything on human-first content in an AI era, and how you can use the same principles to grow your own startup or personal brand with calm, deliberate focus instead of chaos.
Why AI News Felt Like Noise And How He Cut Through It
When generative AI first went mainstream, feeds were full of demos, viral threads, and “magic” prompts. Most people in tech did the same thing:
They saw something blow up, tried it once, said “wow,” and then had no reason to come back.
The problem was not curiosity. It was lack of practical value.
Common issues kept showing up:
- People did not see clear, ongoing benefit to their daily work.
- Tools felt like toys instead of part of a serious workflow.
- The space moved so fast that news felt overwhelming, not helpful.
Rowan saw this pattern early. Instead of trying to scream louder than everyone else, he chose a different frame: no‑noise AI news.
He focused on a simple promise:
Daily AI updates, explained in clear language, with a short take on why it matters to a regular tech worker.
That meant:
- Curating only what was truly important.
- Adding a direct, human opinion on each story.
- Explaining the “so what” for people in real jobs, not just researchers.
At the same time, everyone in tech was talking about “vertical AI SaaS” like AI for law, AI for accounting, and other niche products. Huge AI startups were raising money around that idea.
Almost nobody was talking about verticalized content creators. The person who owns “AI for law content” or “AI for healthcare operators content” can build distribution and trust long before a product exists. That gap is still wide open for new founders and creators.
Rowan chose a broad AI audience, but the same pattern applies: pick a slice of the world, then become the calm, reliable signal in a noisy stream.
From College Dropout To AI Newsletter Founder
Realizing School Was Not His Learning Engine
Before he ran the largest daily AI newsletter, Rowan was a psychology student who could not remember the lecture he had just sat through.
In his words, there was a moment in class when a friend asked a question about the material they had just covered. His friend remembered all of it. He did not. That simple exchange made something click.
He did not learn well by sitting in a room and listening. He learned by building things.
So he started a second education on his own terms:
- Teaching himself to code by building websites.
- Learning design by making real graphics.
- Using “YouTube University” and Twitter to expand what he knew.
- Posting content to see what worked and what did not.
That pattern will feel familiar to many startup founders. Learning through projects, not lectures, tends to show up again when they build companies.
The ChatGPT Spark And The First Big Bet
When ChatGPT was released to the public, Rowan had already been exploring AI for a few months. A day or two after launch, he wrote a simple thread on X (Twitter):
“Here are some cool things I’m exploring in ChatGPT.”
That thread went super viral.
He expected the wave to fade. It did the opposite. Every day, his AI threads performed better. Instead of treating it as a one‑off, he did what strong founders do when they find traction. He went all in.
He dropped out of university and committed to posting a thread every single day about AI. The consistency, paired with timing and genuine curiosity, created compounding attention.
As his audience grew, a new question appeared:
How do you turn viral posts into a real business?
The answer was simple and old‑fashioned: start an email list.
Turning Tweets Into A Newsletter Startup
Rowan launched The Rundown AI as a daily email off the back of his growth on Twitter. Over time it became:
- The largest daily AI newsletter in the world (by his account).
- A daily AI education product, not just a news digest.
- A media startup with over 2 million subscribers and close to 50% open rates.
While many newsletters plateau, The Rundown kept compounding. That reach opened doors that surprise people who only see the end result. When he interviewed Mark Zuckerberg about Llama 3.1 or spoke with Sam Altman, some were confused about how a “newsletter guy” landed those guests.
The answer is straightforward: distribution plus focus.
He was not a journalist, did not attend journalism school, and had dropped out of college. What he did have was a large, engaged audience that cared deeply about AI. That kind of distribution is what many startups pay millions for.
If you want a good comparison of how nimble founders with real distribution can outperform large firms in AI, the breakdown in How startups beat big companies in AI is worth reading alongside Rowan’s story.
Once the newsletter was working, everything else, from interviews to partnerships, built on top of that core.
The Human Edge: Why AI‑Written Newsletters Fall Flat
When The Rundown started, there were already many AI newsletters in the market. A lot of them took what looked like the obvious path.
“If it is an AI newsletter, it should be written by AI.”
That logic sounds clever. It also explains why many of those newsletters disappeared.
Rowan and his team chose the opposite approach:
- They put real faces behind the newsletter.
- They stated clearly that the content was human‑generated.
- They used AI only for support work like structuring and research, not for the actual take.
The reason is simple. AI‑generated content at scale often lacks soul. It reads clean but empty. Most readers can feel that something is off, even if they cannot explain why.
That human layer is not going away anytime soon. Unless we get chips in our brains, people will still care who is speaking to them, what that person stands for, and how they think.
Rowan has a useful analogy for where content is going.
Walmart vs Whole Foods Content
He sees the future of content in two large buckets, with a hollow middle:
- Walmart content: AI‑generated, everywhere, cheap to produce, highly personalized, funny, sticky, and infinite in volume. This will fill most feeds.
- Whole Foods content: Human‑driven, opinionated, attached to a real person, with clear taste and strong trust. This will be rare but powerful.
The top 1 percent of creators who keep deep humanity in their work will probably gain even more influence. They will have more trust, build bigger brands, and distribute ideas or products farther.
Everything bland in the middle gets squeezed.
If you are a creator, founder, or builder inside a startup, the advice is direct: be as human as possible. Tell people who you are. Show how you think. Use AI to help, not to erase your voice.
For a concrete sense of how he structures that human‑first AI content, it is worth looking at The Rundown AI, which shows the exact positioning and promise he delivers every day.
A Simple Roadmap For New Content Creators
Rowan’s system is not mystical. It is structured, repeatable, and especially helpful if you are starting from zero and building a startup or personal brand alongside your main job.
Step 1: Decide If You Are A Talker Or A Writer
The first question is not which platform to use. It is more basic.
Do you enjoy talking or writing more?
Rowan realized he fit the writing archetype, so he started on Twitter, where text and tight ideas perform well. For AI in particular, X (Twitter) is still a core hub where:
- Research breakthroughs are shared first.
- Demo videos appear before they hit blogs.
- Papers and code drops are often released by the authors directly.
If you want to get deep in the weeds of AI, you need some presence there, even if it is small at first.
Step 2: Use A Voice‑To‑Post Workflow If You Hate Writing
Many founders and operators say the same thing:
“I have ideas, but I’m terrible at writing.”
Rowan has a simple workaround that keeps your content human but makes the process faster.
Here is the workflow:
- Install a high‑quality AI transcription app on your phone, such as Whisper‑based tools.
- Go for a long walk and talk into your phone for 1–2 hours. Say “uh,” say “um,” speak in fragments. It does not matter.
- Let the app transcribe your rambles into text. Those raw notes are your real thoughts, not AI ideas.
- Export your past best posts (for example from Twitter), pick 20 or more top performers, and train a Claude project on your writing style.
- Paste your raw transcribed ideas into that project. Let it turn the mess into a draft that sounds like you, then carefully revise.
The result is a post that is roughly “90 percent ready,” rooted in your experience and point of view, without you staring at a blank cursor.
If you want a broader playbook on newsletter building to pair with this, the breakdown in How To Build and Grow a Successful Newsletter in 2025 lines up well with the systems Rowan uses.
Step 3: Respect The Differences Between Platforms
Short‑form content looks similar on the surface across platforms, but the hooks are not identical. Rowan’s team has noticed patterns in how many lines you get to grab attention:
| Platform | Typical hook length |
|---|---|
| 2 lines | |
| 3 lines | |
| Instagram Reels / Shorts | 1 line |
Those are not strict rules, but they are a good starting point. On Twitter you might build a three‑line narrative hook. On Instagram you might need one sharp sentence over a strong visual.
Scaling With AI Avatars Without Losing The Human Core
About a year into running The Rundown, Rowan faced a common founder problem.
Everyone told him he needed to do video. He agreed. But he was working around 80 hours a week as a founder and content creator. There was no extra block of time for daily on‑camera recording.
The team asked a simple question:
How do we turn his writing into short‑form video without needing him on set every time?
Their answer was to clone his face and voice using AI, then keep all the actual ideas and scripts human.
The workflow looked like this:
- Take Rowan’s existing writing and newsletter takes as scripts.
- Use an AI avatar and voice model to “perform” the script.
- Add B‑roll, motion, and tight editing to shape a clear story.
- Publish across Reels, Shorts, and Instagram.
Because the ideas, opinions, and structure were still his, the content felt aligned with the newsletter, even if he was not sitting in front of a camera every day.
The result was strong: the Instagram account passed 160,000 followers in about a year, powered heavily by this avatar‑based workflow.
That model is now becoming more common among serious creators. You can see similar patterns in how some AI‑focused creators and founders working on tools like Palo, an AI startup helping creators make viral content, think about scale.
Using AI To Draft, You To Decide: Ads, Copy, And Beyond
The same Claude‑based approach he uses for tweets also extends into the business side.
For example, when The Rundown needs new ad copy for partners, the team:
- Trains a Claude project on their best‑performing ad creatives.
- Feeds in details about a new sponsor or campaign.
- Lets Claude generate a draft, then edits it heavily by hand.
AI handles the first 60–80 percent, but the team still decides what they are comfortable putting their name on. That combination scales output without diluting the brand.
If you want to see another operator’s take on growing an AI newsletter, How I'm Going To Grow My AI Newsletter is a useful companion read and shows how similar these systems look when you strip away personal brand differences.
Three Daily Habits To Build Real AI Intuition
Rowan often repeats a simple line: AI is taught through habits. Not through a one‑time course, but through what you do every day at work.
He offers three habits that, if taken seriously, can put you in the early‑adopter tier inside your company or startup.
Habit 1: Ask “Could AI Do This?” Every Day
Any time you do something repetitive or tedious, ask a single question:
“Could AI do this task?”
If you are not sure, you can literally paste your task into a chatbot and ask, “Could AI do this X task?” That one question shifts your mind into scanning for automation opportunities instead of accepting everything as fixed.
A simple trick is to put that question on a sticky note on your laptop. Over time, you will build a mental map of what AI can and cannot do well for your specific role.
Habit 2: Set Time To Test Tools, And Accept Failure
Once you spot a task that looks automatable, pick a tool and test it. Rowan likes to use Sundays for this, but any weekly slot works.
Sometimes you will succeed and free up real time for higher‑level work. Sometimes you will fail and learn where AI still breaks or falls short.
Both outcomes are useful. Together they build what he calls AI intuition: knowing when a tool is better than you and when it is not.
If you want more structured ideas to test, the free HubSpot for Startups resource, 100+ Creative AI Use Cases Beyond Cold Outreach, is a strong catalog. It includes over 100 real examples across 30 industries, with clear descriptions, step‑by‑step guidance, expected business outcomes, and difficulty ratings. It is the kind of reference that can spark startup ideas or internal projects for months.
Habit 3: Share What You Learn
The last habit is simple and underrated: share your learnings.
Sharing can mean:
- Posting threads or LinkedIn updates about what worked.
- Dropping short summaries in your team Slack.
- Explaining a workflow to a friend or family member.
Teaching forces clarity. It also quietly turns you into “the AI person” around your office or network, which tends to open up career paths, partnerships, and startup ideas that are hard to predict in advance.
If you want to pair Rowan’s experience with a broader view of how vertical AI companies are being built, pieces like Vertical AI Explained: The Next Generation of Tech Titans and Top 15 Vertical AI Companies & Startups show where the product side is going while his story shows how media and distribution keep up.
Why Creators Might Be The Next Billionaires
Rowan is not very active in politics, but he pays attention to influence. He points to people like Joe Rogan as proof that content creators can shape public opinion at national scale.
Fifty years ago, if you told a farmer there would be a job called “content creator” who could influence an election, they would probably laugh. Today, that is normal.
Rowan thinks that influence will only get stronger, even as it gets harder to reach that level. In a world filled with generated content, the creators who build deep trust and hold real attention will sit in a powerful spot.
Content creators, especially those who think like startup founders, have three big assets:
- Distribution: direct reach to a defined audience.
- Data: a live feedback loop on what the audience cares about.
- Optionality: the ability to build products, services, or companies on top of that audience.
Everyone talks about vertical AI startups like “AI for law” or “AI for accounting”. Very few people talk about vertical content creators like “AI for lawyers content” or “AI for CFOs content”.
That area is still blue ocean. Almost nobody owns “AI for law content” at scale yet. Somebody will.
If your goal is to build a startup around AI, starting as the clear voice for a specific group of professionals is one of the most efficient ways to stack trust before you ship anything.
The Age Of The Idea Person
Near the end of a conversation with Sam Altman, Rowan came away struck by one idea: the future might include zero‑person companies, where even the CEO role is handled by an AI agent.
You do not have to agree with that vision to feel the pattern. Each year, more of the execution work can be handled by tools. That makes the quality of ideas, taste, and direction even more valuable.
We are in a period where, if you have a thoughtful idea and are willing to put in uncomfortable hours, you can assemble a serious AI‑powered startup faster than at any time before. Distribution, through content and newsletters, is one of the most durable force multipliers in that process.
Rowan’s story is not just about luck. Yes, there was timing. But there was also a clear pattern:
- He noticed an information gap.
- He went all in when he found traction.
- He kept the content deeply human while using AI as a force multiplier.
- He treated his newsletter like a real startup, not a hobby project.
If you are building something similar, those are the levers to study.
Conclusion
The rise of The Rundown AI shows what happens when a founder treats content as a product, not an afterthought. You do not need a degree, a press badge, or a giant team to build a serious AI newsletter startup. You do need clear focus, human‑first opinions, and habits that keep you close to how AI actually works in practice.
Start by asking where the noise is loudest and where you can offer calm, useful signal. Decide whether you are a talker or a writer, build a lightweight system around that strength, and use AI to support, not replace, your voice. From there, your job is to keep showing up, ride your lucky breaks, and treat your audience with the same respect you would treat your earliest customers.
The tools will keep changing. The founders and creators who combine human depth with disciplined systems will keep winning.
0 Comments