The Code/X ArchiveView on X
Eliot Horowitz

@eliothorowitz

Open-source is essential. I built MongoDB, one of the world’s most popular open-source databases. Now I’m building Viam, an open-source harness for physical AI.

I don’t want a world where millions of developers are dependent on the paywalled technology of just a handful of companies. No one should.

I want to see 1B+ new physical AI deployments over the next ten years.

Here’s why open source is essential to getting there. 🧵
954143.3K1.2K
Eliot Horowitz

@eliothorowitz

First off — your model is not your moat.

The robotics data everyone trains on is public, small and shared (Open X-Embodiment, DROID).

That means models trained on it converge.

Your moat comes from data collected on your own machines, and from what you do with the models.
3311514
Eliot Horowitz

@eliothorowitz

Moat =
1. Proprietary data. The data from your machines, in your environment, doing your tasks.
2. Know-how. Knowing how to use that data, and how to break the problem down correctly.
3. Creativity and grit. No matter how good AI is, building a product and a business is hard. You have to make it happen.
1810320
Eliot Horowitz

@eliothorowitz

That’s your IP. You should own it.

Everything else should be open.

Open models, open weights, open coding systems.
12745
Eliot Horowitz

@eliothorowitz

Time and time again we see explosive innovation happen in open ecosystems, not in closed ones.

For physical AI and robotics to truly take off it needs democratization, not consolidation.

Wisdom of crowds > the blessing of a few companies that want to own the future exclusively.
15635
Eliot Horowitz

@eliothorowitz

What does taking off look like?

Millions of developers building tens of thousands of startups

…all distributed across a vast expanse of the economy and physical world.
13462
Eliot Horowitz

@eliothorowitz

Physical AI and robotics development become part of every business.

Small machine shops to enormous e-commerce companies.

Global shipping behemoths to a local commercial laundry.

And every conceivable thing in between with a need.
22422
Eliot Horowitz

@eliothorowitz

Addressing some obvious critiques:

1. “Scale wins and whoever has the most data trains the best model.”

The data that matters is spread across millions of deployments, in factories, farms, and warehouses that no model can see. That argues for open infrastructure owned by the people who have the data.
13421
Eliot Horowitz

@eliothorowitz

2. “Open weights already exist.”

Sure, but weights are a start. The tooling, data pipelines, and deployment layer have to be open too, or the weights do little for the people who need them.
13404
Eliot Horowitz

@eliothorowitz

We all know that one of the biggest proponents of open source just exited the main stage.

I know from experience — business and open source are not opposed.

Let's make open source the standard in physical AI and robotics.

Full thoughts here.

eliothorowitz.com/blog/2026/11/1…

eliothorowitz.com

Open Is How Robotics Wins · Eliot's Blog

456217
End of thread