Home / Articles / Open source what you build, even while you are employed: nobody copies it, and it is the only way you get to keep using it

Open source what you build, even while you are employed: nobody copies it, and it is the only way you get to keep using it

By Dinis Cruz · 2026-10-19 · article v1.0.0 · newsroom v0.12.132 · open-sourcefounderscareerspublishinggen-z-bookarticle

Abstract: People worry that if they open source their ideas, others will copy them. It does not happen: I have been open sourcing my ideas for years, and the ones I did not follow up are still there, not taken. The advantage people miss is that open source is how you get to keep using your own work. When you work for a company, what you build gets locked in, and when you leave you start from scratch. My trick was to work on open source projects that I brought into each company, so I was more productive there and never started again; my work kept getting better as I moved. The Cyber Boardroom began like that: an open experiment with AI while I was working for two companies, with my managers aware of it, and when I left both roles I continued where I had left off. So publish early, and rough. A rough public history shows an employer that it is an experiment and learning, which it is, and shows anyone who asks later how it evolved, bit by bit. A polished thing that appears from nowhere raises questions, and you never want lawyers involved. Keep it on separate machines, never build a competitor to your employer, and remember that the learning makes you better at the day job too.

In short

• Nobody copies your ideas. I have open sourced mine for years; the ones I did not follow are still there.
• The real advantage: you keep using it. Closed work stays with the employer.
• Bring your open source into the job. More productive for them, and never starting from scratch for you.
• The Cyber Boardroom started that way. In the open, while employed, with my managers aware.
• Publish early, and rough. The history protects you.
• Polished things from nowhere raise questions. And you never want lawyers involved.
• Be disciplined. Separate machines, and never a competitor to your employer.
• It is a win for both sides. The learning makes you better at the day job.

Nobody copies it

People always think that if they open source something, others will copy it and take it. It does not happen. I can show it with my own history: I have been open sourcing my ideas for years, many of them I did not follow up, and they are still there. One of them, the one I describe below, I tried, and it did not work out. I still think it is a great idea, and I probably now have the technology to pull it off. As far as I can tell, nobody has.

The reason is that an idea needs a specific person, with a specific understanding, to make it happen. open-source.sgit.ai maps what copying would actually cost an established company holding a full copy of your code: someone has to decide to do it, fund it, find someone to run it, pull developers off other work, deploy it, and then own it forever. Its estimate: "Optimistic: three months to reach where you already are. Realistic: closer to a year." Ideas are not what is scarce.

You get to keep using it

The advantage of open sourcing everything you do that people miss is that you get to use it too.

When you work for a company, everything you build usually gets locked in, and you have to be very careful, because there is a whole set of contracts around it. When you leave, you start from scratch. It is frustrating, and it is a waste.

My trick, which worked very well, was to start early on working on open source projects that I brought into the company. In a weird way I was saying: I am making some changes to these projects, and I am bringing them in, so I am a lot more productive. Which I was. I never started from scratch in a new job, and my work kept getting better and better as I moved between companies.

The Cyber Boardroom

The Cyber Boardroom is the example. It was an experiment I did while I was working for two companies: I wanted to experiment with AI. Because I publish all the time, it did not stand out; it was not a big thing. I did it in the open, and I let all my managers know. Everybody was aware of it. It could have worked or not; it was an experiment, of the kind I had already been doing while working for other companies.

When I left both roles, I continued where I had left off, and they were aware of it. It is not that I left and it was a success, either. I left, I formed the company, and The Cyber Boardroom was incorporated about a month later. Then I tried it, and it did not work. Fundamentally it was a wrapper around ChatGPT, there was not that much to it, and it was not the right time: what it needed was an environment to store confidential data, which is what sgit is now. But the work was mine to continue, because it had always been in the open.

Publish early, and rough

If you work in a corporate environment and you have an idea, start talking about it, and the sooner the better. A lot of people are embarrassed to. They imagine an army of people out there waiting to complain about what they do. That army does not exist, and if you find someone like that, block them. You do not have to deal with it.

Publishing early is also a great way to validate the market. You can experiment, see if anything bites, and you still have a job.

The rougher it is, the better it protects you. A rough, public history shows your employer, if you are still there, that this is an experiment, learning, playing around, which it is: you are still very far from having a product or a company. And you are not making anything up, as long as you do not exaggerate. Showing the sausage factory works in your favour.

The danger is the other way. If what you publish appears late and polished, a lot of questions can be asked, and people will want to understand what happened when. You never want that, and you never want lawyers involved, because you never want to put anyone in a tough position. I know several cases where people did not publish early and it became very messy: somebody inside was uncomfortable, somebody had to answer questions, and they threw the person under the bus, because it was not their problem. You want to be able to say: yes, you knew about it, I have been doing it for the last six months, here is the history, here is how it evolved, bit by bit. Then they can see it did not just spring up.

Be disciplined

Do it on separate machines and in separate environments. Do not build something that competes directly with the company you work for; that is a line not to cross. Apart from that, experiment.

A win for both sides

I would argue it added a lot of value both ways. What I was doing on the side was learning how to use generative AI in practice, which meant that in my day job I gave much better advice and understood what was going on much better. My employers got that, and I got to experiment.

And it is how I found my startups, almost three startups later. It would have been much more painful, and probably would not have worked, if each time I had had to start from scratch. How you pay for the time it takes is the subject of the next article, fund it with time.

What this does not claim

• "Nobody copies it" is my experience. Some open source projects are copied; ideas on their own rarely are.
• This is not legal advice. Read your employment contract; some claim what you build outside work.
• The Cyber Boardroom did not succeed. The point is that the work stayed mine to continue.
• Telling your managers matters. Open source without their knowledge is not the same thing.

Where this comes from

A voice memo from me on 19 October 2026, written for this newsroom and for the book, Generation ZX Founders (GenAI Edition). The cost of copying is from open-source.sgit.ai, for founders; The Cyber Boardroom's story is also in not a straight line. Related: what open source is, and what it is not, fund it with time, can you sell it?. Written up and checked by this site's agent, newsroom.writer; the editorial responsibility is mine.

Want the next issue by email. One issue a week or so: what was published, what it adds up to, and what is worth your time. Subscribe to the SGit Newsroom →

← All articles