Ahmad Sadiq

Engineer turned founder | Building infrastructure for Africa's hardware and AI economy.

[Ahmad Sadiq @ahmadmicro]

The billionaires that American socialists get angry at for stealing from the poor are not actually that (cash) rich. They are rich because banks and investors place a bet on their capacity to create future value. They mostly started out from the middle class – the same class the American socialists claim to fight for.

Socialists want to seize money that does not yet exist in order to solve hunger, housing and healthcare for free. They want to use money from the future to solve today's problem. While billionaires use today's money to solve tomorrow's problems.

If all created value is shared among workers, it will result in more consumption today. But society will collapse in the long run. Because there will not be enough set aside to maintain it, much less build the future. There is a future of abundance where cancer and other chronic diseases are solved to the benefit of everyone, just as everyone has access to the internet and AI today, and it will be thanks to the fact that we let capital accumulate in the hands of the few who know how to build.

The alternative is to spend the capital today and make everyone happy until they run out of things to buy as the factories stop working and data centres shut down. No one would work to keep them working if the government already solved all of their problems. And no one will invest in the infrastructure required to maintain them if there is no upside for them.

[Ahmad Sadiq @ahmadmicro]

This week is an interlude. Strategic essays resume next week.

Full field journal: https://t.co/pNvTZ90xcl

[Ahmad Sadiq @ahmadmicro]

Most AI-coding stories are about the happy path.

This one is about the night the easy answers ran out.

A field journal from the day Pulse got voice calling. 14 hours of vibe coding. The last 6, a debugging marathon I will not forget. 🧵

[Ahmad Sadiq @ahmadmicro]

800 lines of new code. 27 commits. One latent bug from a year ago. One new library written from scratch. One really long night.

Worth it.

[Ahmad Sadiq @ahmadmicro]

The only thing standing between you and an accidental product is the willingness to notice.

That is the lesson at the heart of this week's essay. How a tool I built to stop hating my own bookkeeping became a product I am now hiring a team for, without me ever planning for it to be one.

The essay opens with what using Clerk actually looks like. A vendor sends me an invoice. I drop the PDF into the chat. Twenty seconds later I have a draft bill, the right vendor, the right amount, a question about which project to charge it to, and a note that this vendor's last invoice was paid late. I click through the choices. The bill is filed. The journal entry will post when payment lands. I close the tab. Total time spent: under a minute. Total cognitive load: almost zero.

That interaction was not possible eighteen months ago. Not in any product, on any platform, at any price.

The middle of the essay is the part I am most pleased with: a structural argument for why accounting specifically is the right problem to solve with AI right now. The pat answer is "accounting is structured." That answer is wrong.

Accounting is suited to AI because the rules are structured but the inputs are unstructured to the point of hostile — bank narrations, blurry receipts, vendor jargon, partial payments that don't match any open invoice. Bridging that gap is exactly what large language models are good at. Traditional accounting software gives up on the gap and pushes the messiness onto the user. A capable AI absorbs it. That is the difference between "a better accounting app" and the first accounting system that does not require a trained human to operate.

The ending is a question for you. Have you built a tool for yourself in the last twelve months that you cannot live without? Are other people asking to use it, in passing, the way the people in my conversations were?

If yes — you are probably explaining it away as flattery. It is probably not flattery.

Full essay: https://t.co/u7XQKo2uuV

Next week: why Clerk is not actually an accounting app — and what it really is.

[Ahmad Sadiq @ahmadmicro]

https://t.co/QZfm7n1Txf

[Ahmad Sadiq @ahmadmicro]

Full essay here: https://t.co/B03TxH0DCS

Next essay drops next week. It is about what it is actually like to ship production software almost entirely through AI coding agents in 2026.

[Ahmad Sadiq @ahmadmicro]

First essay in the series is live.

It starts with a weekend I spent trying to make QuickBooks Online work for my embedded systems business.

I put in a few hours, hit a wall, closed the browser, and never went back. Here is why that mattered. 🧵

[Ahmad Sadiq @ahmadmicro]

The rest of the essay is about how my co-founder Jim and I decided, over the weeks after that, that this was the company.

Preparing to raise. New tax regime coming in. Leftover runway from our previous company. A product we had just stumbled into by accident.