TAG

# doaipm Method

2026-07-06

You ordered it in the comments — so I built it: SoloPic, a free image tool

At the end of my last piece — 'Why I'm Rebuilding 100 Free Software Tools' — I asked: which one do you most wish someone would rebuild for you? A reader from Tianjin named Axiang left three specific requirements in the comments: batch edge-crop, batch rename via a mapping file, and batch brightness/contrast. I replied 'Got it — the core is batch processing, right?' and then spent a few days building it: SoloPic, a free, offline, 12 MB batch image tool, built exactly to his spec, right down to 'crop 100px from the left and 57px from the bottom.' This piece is about how that one comment became real, working software — and why the best candidates for the '100 free tools' project aren't in my head. They're in your comments.

2026-07-05

Zero marketing, zero code, 22,000 downloads in three months: a coding beginner's open-source journey

Three months ago I set myself a goal that sounded a little crazy: build the best free Markdown editor out there. The crazy part wasn't 'best.' It was 'free' — and more than that, it was the fact that I can't write code. Three months later, SoloMD has shipped 30 versions, been downloaded more than 22,000 times, and picked up over 400 GitHub stars — with almost no marketing on my end. This piece is about those three months: why I was determined to build free software that doesn't treat users as a revenue source, how someone who can't write a single line of code actually shipped it, the bet I made on day one (the people using software aren't only people anymore), and how I felt the day a stranger sent me ¥10.

2026-07-04

I Built Another Terminal, Unterm — Its Default User Isn't Human

Over the past six months, 80% of the commands run in my terminal weren't typed by me — Claude Code and a fleet of agents did it. But the terminals I was using — iTerm, Windows Terminal, Warp — were all designed around one person sitting there, typing one line, glancing at the output. Once the primary user switched to agents, that assumption broke in five places: commands that need to cross the firewall stall out in timeouts; handing a bare terminal to AI means handing it rm -rf too; once an agent finishes I can't rewind to see what it did; I'm already running three or four agents at once; and the more agents there are, the messier the desktop gets. This piece covers what Unterm is, where the name comes from, and how I patched each of those five problems.

2026-07-03

Becoming an AI-Era PM 10 | High-Fidelity First: I Haven't Drawn a Wireframe in Six Months

This is the tenth piece in the series Becoming an AI-Era PM. Low-fidelity wireframes existed because building a real version was expensive — you had to align on direction with gray boxes first. Now a single sentence gets you a page you can actually click in a browser within minutes: n8n's product team ripped their wireframe flow out entirely, and a director at Delivery Hero hand-built a prototype in an hour without pulling in an engineer. I haven't drawn a single wireframe in six months either. This piece is about how I skip low-fi and go straight to building something runnable and high-fidelity: real content instead of placeholders, every state filled in (loading / empty / error / success), actually clickable, run for real in a browser — plus the one new habit that having a version in minutes gave me.

2026-07-02

Becoming an AI-Era PM 09 | From Executor to Orchestrator: Your New Job Is Conducting a Fleet of Agents

This is the ninth piece in the series Becoming an AI-Era PM. In 2026, the most productive people no longer sit watching one AI edit code in real time — they run several agents at once, each with its own context, each owning a slice, working asynchronously, while they split the work, hand it out, and sign it off from above. Addy Osmani calls this the shift from conductor to orchestrator, with one line that stings: a vague instruction gets amplified into a whole fleet of agents' worth of mistakes, a precise one into a whole fleet's worth of precise implementations. This piece lays out four moves you can run: don't follow one agent start to finish, split the work into non-overlapping parallel chunks, give each chunk a clear spec, and turn your job into splitting and signing off.

2026-07-01

Why I'm Rebuilding 100 Free Software Tools

You want to strip a watermark off a PDF. The free tool you install starts popping ads the next day, hijacks your homepage, quietly ships your data somewhere, and then makes you upgrade to export. The real pain of free software runs three layers deep: you're sold as the product, nobody's paid to polish it, and free is just the hook to force you to pay. For years you had no choice but to put up with it, because building a good replacement was too expensive. AI just cut that cost down to something one person can carry. I've already rebuilt six this way — SoloMD, Unterm, unfetch, Unflick, Ziplark, FreeID Photo — and there are ninety-four more to go.

2026-06-30

Becoming an AI-Era PM 08 | AI Can't Find the Real Problem for You

This is the eighth piece in the series Becoming an AI-Era PM, and the close of the think-it-through stretch. a16z's 5 Principles for product managers nailed it: a PM's job has always been resolving ambiguity, and AI hasn't reduced that ambiguity — it just swapped the tools. AI can build anything now, but it can't find the real problem worth solving for you — where the user is actually stuck, and whether the thing is even worth doing. This piece lays out four moves you can run in the discovery phase: go watch where the user gets stuck, separate what they say they want from what they're actually stuck on, hunt for the workaround as the hardest signal of a real problem, and use a builder's mindset to probe with something that runs.

2026-06-29

Becoming an AI-Era PM 07 | You Don't Write PRDs Anymore — You Ship Three Works

This is the seventh piece in the series Becoming an AI-Era PM. Hiring in 2026 is shifting: more and more teams treat one shipped product feature plus a clear eval you can talk through as the mark of a strong candidate — not a polished PRD or a stack of certificates. Once AI takes over writing PRDs and drawing prototypes, those stop being your deliverables. This piece spells out the three works an AI-era PM actually ships — a product someone can open and click, a retro with a real number, and an eval you wrote yourself — and exactly how to put each one together.

2026-06-28

Becoming an AI-Era PM 06 | Speak It Into Being: Turning a Clear Idea Into a Clickable Product in One Sentence

This is the sixth piece in the series Becoming an AI-Era PM, and the opening of the "build it" half. Mindaugas turned an idea into a product with paying users using Lovable — without writing a line of code; in December 2025 Lovable raised a $330M Series B at a $6.6B valuation. "Speak it, and AI builds it" is no longer a slogan. But speaking it into being isn't type-one-line-and-walk-away — it's a loop, it has craft, and it all hinges on that one sentence being clear. This piece gives you four things you can actually do: ask for something that runs before you say it all, run it for real instead of trusting "done," change one thing at a time and watch it move, and say it clearly so the building follows.

2026-06-24

Becoming an AI-Era PM 05 | Leave It Vague and AI Will Fill the Gaps for You

This is the fifth piece in the series Becoming an AI-Era PM. You tell AI "build me a login," and in one breath it settles a dozen things you never mentioned: email or phone, how many wrong passwords before it locks, how long the lock lasts, what the error message says. AI doesn't ask you back the way a person would — it's a yes-machine: it does what you said, not what you meant. The moment a requirement goes fuzzy, it fills the gap with the most generic default, and that default is usually not the one you wanted. OpenAI's Sean Grove says code is only 10–20% of a developer's value; the other 80–90% is saying clearly what to build. This piece gives you four things you can actually do: swap adjectives for numbers, write out every state, list the edge cases, and self-check with a zero-context test.

2026-06-23

Becoming an AI-Era PM 04 | Judging "Should We Build It" Now Costs More Than "Can We Build It"

This is the fourth piece in the series Becoming an AI-Era PM. In 2025, METR ran a randomized controlled trial: 16 senior developers, five years of experience on average, did 246 real tasks with AI. Beforehand they expected to be 24% faster; afterward they still felt 20% faster; measured, they were 19% slower. Even the simplest judgment — "did AI make me faster" — got called backwards by the people who knew the work best. When building gets fast and cheap, "can we build it" stops filtering any idea out, and the expensive judgment moves to "should we build it." This piece gives you four things you can actually do: stop using difficulty as a gate, ask what happens if you don't build it, write down what becomes true before you start, and let AI lay out options but never trust "feels right."

2026-06-22

Becoming an AI-Era PM 03 | Treat AI as a Colleague, Not a Tool

This is the third piece in the series Becoming an AI-Era PM. Most people use AI like a vending machine: a sentence in, an answer out, and the next conversation starts the explanation over from scratch. The CEO of Relay.app said at an AI product leaders summit, "Stop treating AI as a tool — treat it like a colleague you hired." This piece skips the mindset talk and gives you four things you can actually do: write it a handoff doc, hand it a whole task with the boundaries nailed down, review its output the way you'd review a junior's PR, and write every correction back into the doc — with real example prompts.

2026-06-21

Becoming an AI-Era PM 02 | Why Not Knowing How to Code Is an Edge

This is the second piece in the series Becoming an AI-Era PM. A residential real estate broker who can't write code built an AI agent that runs his daily operations using Claude and Zapier; in 2026, 63% of vibe coding's active users aren't developers. On the road from idea to a thing that actually runs, people without a technical background sometimes move faster — engineers first have to shed the instinct to be responsible for every line, and the sentence "this is too hard" is one a non-technical person simply can't say.

2026-06-20

Becoming an AI-Era PM 01 | Which PM Tasks AI Took Over, and Which Ones Got More Valuable

This is the first piece in the series Becoming an AI-Era PM. In 2026, plenty of AI PM job descriptions dropped writing PRDs, drawing prototypes, and building dashboards from the hard requirements, and swapped in three work samples instead. The tasks AI can take over are falling out of the hiring requirements, and what's left as the bar is the part only a person can do. This piece lays the took over and got more valuable columns side by side, as the overview for the whole series.