The ship and the speedboat

2026-10-15

I think hierarchy and power in software are decentralizing. For a long time code was the expensive part. Building a serious product took an enormous number of person-hours, and the companies that could pay for those hours, and organize them with process, were the strong ones. The code itself was the moat.

That is changing. Code is now cheap compared to what it used to cost. Expertise is also more spread out, instead of sitting in a few central places. It is no longer only the people with the most experience, or with the biggest teams behind them, who hold the power. My main point is simple: being adaptable now beats being big and strong.

One developer, seven apps

The example I keep coming back to is ArtCraft. It's an open-source suite by a developer named Brandon Thomas, who has about 15 years of programming experience and was previously CEO of Storyteller.ai. He was fed up with Adobe's cancellation fees, so he built seven apps that recreate Photoshop, Illustrator, Premiere, Lightroom, After Effects, InDesign and Acrobat Pro. It's written in Rust, it's MIT/Apache licensed, and he built it with help from AI, Claude in his case.

To be fair, it's early and rough. Reviewers report glitches, and he first promised full feature parity, then said it would take months. Adobe still has real strengths: professional workflows, plug-ins, color management and long-term polish. Even so, HSBC downgraded Adobe to Hold this week, citing exactly this kind of AI clone risk, while also warning people against downloading ArtCraft for security reasons.

The point for me is not that ArtCraft already replaces Adobe. It doesn't. The point is that one developer can now build something in that direction at all. Not long ago, I think even one of those apps would have needed a large team and years of work. Now one person is working on all seven, rough as they are, and I think the gap between rough and good will close a lot faster than people expect. For the first time it feels realistic to me that one person can take a shot at a big company, because writing all that code is no longer what stops you.

The ship and the speedboat

I picture big companies as a huge ship. It's stable, heavy and strong, it carries a lot and it doesn't tip over in a storm. But turning it takes forever. Every change of direction has to go through process, hierarchy and approvals, and all that bureaucracy slows it down.

Individuals and small teams are speedboats. Next to the ship a speedboat is small and fragile, but it can turn almost immediately. When the direction everyone has to go in changes, the speedboat is already heading there while the ship is still discussing it. Right now the direction has changed. Building software looks very different than it did a few years ago, and I think big companies are going to be really slow compared to the rest of the world. For some of them I think it will hurt badly.

So in my view, adaptability beats strength. Size used to protect you. Now it can be the thing that holds you in place.

What still protects the ship

I don't think big companies are defenseless. What still protects them is mostly not their code. It's trust, existing customers and distribution. People already pay them, already rely on them and already know where to find them. A new tool has to earn all of that from zero, and that takes time. A big company can also cut prices or optimize hard when it feels the pressure. These are real advantages, just a different kind than "nobody else can afford to write all this code".

The rules changed

Czech voucher privatization in the 90s is the closest parallel I can think of. Suddenly ordinary people could get a stake in something big, if they understood what was going on early enough. It was far from fair, and plenty of people got burned by funds that were looted. But the lesson I take from it is that when the rules of the game change, the people who pay attention first move first.

This is one of those moments. Today the entry ticket is a laptop and an AI subscription. That's not much money a month, and with it one person can build almost anything, and not only code: graphics, photos and product videos too. I think this window will not stay open forever. At some point it will close, the new players will settle in and a new hierarchy will form. That's a prediction, not a fact. But right now I think the window is open, and it's worth using.

Be a generalist

The practical conclusion I draw for myself is to be a generalist, not a narrow specialist. That doesn't mean staying shallow. When something needs depth, go deep into it. Just don't lock yourself into one narrow thing and build your whole identity on it.

The people I think are most exposed are the ones whose entire status was built on one narrow skill or one career ladder. If that skill gets cheap, or the ladder stops meaning much, the status falls apart like a house of cards. Maybe not yet, but I think it will. Someone who has had blinders on for years and can only do their one thing will have a hard time when that one thing stops being rare.

A generalist has a better chance. You can see where the direction is changing, pick up whatever the new thing needs and turn with it. That's the speedboat again. Right now, being able to adapt is worth more than being the strongest at one thing.

- mind dump from an evening walk in San Francisco