The Higher Ground Strategy
Last week I wrote about making 6,048 GitHub contributions with nothing to show for it. Twelve followers. No revenue. A year of building in private while Lovable generates 100,000 new projects a day.
The response was useful. Several people said some version of: "I'm in the same boat." One person asked the question I'd been avoiding: "So what are you actually going to do about it?"
This is my answer. Not a polished strategy deck. More like thinking out loud with a commitment attached.
The race I've been running
I've been trying to compete on volume. Build more products. Ship faster. Cover more ground. The logic felt sound: if the odds of any single product succeeding are low, build fifty and let the power law sort it out.
The problem with that logic in 2026 is that everyone else can now build fifty products too. A non-technical founder with a Lovable subscription can ship a prototype in an afternoon. The barrier to creating software has collapsed. What used to take me weeks of careful architecture, someone can now prompt into existence before lunch.
I can't outbuild that. Nobody can outbuild a hundred thousand new projects a day. Trying to win on volume against an infinite content machine is like trying to outswim a river. You'll drown tired.
What the survivors have in common
I've been studying the indie hackers who actually make money, not the ones who post screenshots of Stripe dashboards they'll never see again. The pattern is boringly consistent.
Caleb Porzio built Livewire as open source, gave the code away free, and now makes five figures a month from courses, a premium component library, and years of trust within the Laravel community. Someone on Indie Hackers nailed the insight: the code was never the moat. The ecosystem around it was. The trust, the conference talks, the consistent shipping. That is what you can't replicate in an afternoon with a prompt.
A Squarespace plugin developer called Beyondspace grew to 3,600 installs by doing something almost aggressively unsexy: answering support tickets fast, asking for reviews, building a 3,000-person email list with 45% open rates. No growth hacks. No viral loops. Just relentless usefulness in a specific community until word of mouth kicked in.
The pattern is: audience first, product second. Or more precisely: trust first, then build for the people who trust you.
I've been doing it backwards.
Why I can't just "do marketing"
The standard advice for builders who don't distribute is: learn marketing. Start posting on X. Write a newsletter. Do SEO. Run ads.
That advice is technically correct and practically useless for someone like me. I have ADHD, a full-time job, and a pathological aversion to self-promotion that I've spent forty-odd years failing to fix. Telling me to "just start marketing" is like telling someone with vertigo to "just climb the ladder." The ladder isn't the problem.
I need a strategy that works with my wiring, not against it. Specifically, I need an approach where the distribution is built into the product itself, not bolted on as an afterthought I'll inevitably neglect.
The higher ground
Here's what I've landed on. Stop trying to swim in the river. Climb above it.
The vibe-coded pile has specific, predictable weaknesses. AI-generated code contains 2.74 times more security vulnerabilities than human-written code. It accumulates technical debt three times faster. It can't handle its first dependency update without breaking. The founders who shipped it can't debug it because they never understood it.
These aren't theoretical concerns. They're already happening. Forrester predicts 75% of enterprises will face serious technical debt from AI-driven development this year. The "vibe coding hangover" isn't coming. It's here.
That hangover creates a market. Not for more products, but for products that demonstrably don't rot. Products with transparent methodology. Products whose architecture you can inspect. Products built by someone who understands what they've built and will be around to maintain it.
That's my higher ground. Not "I built this with AI too, but mine is better." Instead: "I built this by hand, I can show you how it works, and I'll still be here when you need it fixed."
Three moves
Concretely, this means three changes.
Open source as distribution. My GEO benchmark framework already has 51 assessment dimensions, 128 evaluation prompts, and a six-model AI consensus panel. It's on GitHub. Almost nobody knows it exists. The code itself should be my marketing. Every open issue, every PR, every contributor who finds it useful becomes distribution I don't have to manufacture. The README is my landing page. I need to treat it that way.
Research from the indie hacker community keeps surfacing the same finding: open source creates organic distribution that you can't buy. Developers share tools they use. Blog posts get written about projects without anyone asking. GitHub stars compound. The catch is that you need a minimum audience to get the flywheel spinning. Which brings me to the second move.
One community, deep. I've been spreading myself across solopreneurs, AI practitioners, GEO specialists, and awareness teachers. That's four audiences for one person with a day job. The research is clear on this: a tight audience of 5,000 outperforms a broad audience of 100,000. I need to pick one community and become genuinely useful to it. For me, that's the GEO/AEO space. It's where my products are strongest, where my methodology is most rigorous, and where the market is young enough that trust hasn't been captured yet.
Proof over promises. The vibe-coded pile is all promises. Beautiful landing pages, impressive demos, and then silence when you try to use the thing in production. My competitive advantage isn't that I'm a better coder (though 6,048 contributions suggests I'm at least a persistent one). It's that I can show my working. Open methodology. Published benchmarks. Reproducible results. In a market drowning in AI-generated noise, transparency is the signal.
What this costs me
Honesty compels me to say what this strategy costs. It costs me speed. I can't open-source everything, serve one community deeply, and build new products at the pace I've been going. Something has to give, and it's the portfolio approach. Fifty products by 2030 was the plan. The plan was wrong. Three products that people actually use and trust is worth more than fifty that nobody finds.
It also costs me the comfort of building in private. Opening up the methodology means people can criticise it. Publishing benchmarks means someone might find a flaw. Writing about the journey means admitting that after a year of extraordinary effort, I have twelve followers and no revenue.
But hiding hasn't worked. I tried that for a year. I have the GitHub profile to prove it and nothing else.
The actual commitment
Starting this week:
I'm picking AISearchArena as the one product. The benchmark framework goes fully open. The methodology gets documented publicly. The monthly reports get distributed to every GEO practitioner I can find.
I'm writing for that community specifically. Not "content about AI" for a general audience. Specific, useful analysis of how search engines are changing and what it means for people whose visibility depends on it.
I'm treating my GitHub profile, all 6,048 contributions of it, as a credential, not a scoreboard. The contributions prove I can build. The next twelve months need to prove I can connect what I build to people who need it.
The higher ground isn't a metaphor. It's a position. Above the flood of AI-generated noise, there's a place where quality, transparency, and trust still matter. Getting there requires letting go of the volume game and doing fewer things, better, in public.
I don't know if it'll work. But I know that what I've been doing doesn't. That is arithmetic.
Jamie Watters builds AI-native products at jamiewatters.work. Find him on LinkedIn, X, and WIP.