I stopped comparing OpenGradient to other AI platforms after a few sessions.
Not because the models felt dramatically different. Most AI products are getting good enough that the quality gap is shrinking. What stood out was something less obvious.
I intentionally ran the same prompts multiple times over a few days. Some were harmless. Some contained information I normally wouldn't paste into a public AI tool. The responses weren't perfect every time, but that wasn't what I was watching.
I was paying attention to whether I trusted the environment.
The numbers made the question harder to ignore. OpenGradient recently raised $9.5 million and has already processed more than 156,000 private inferences through its network. That's not experimental traffic anymore. People are actually using it.
What's interesting is that another model with slightly better benchmark scores wouldn't change my behavior much. A system that gives me confidence about where my data goes probably would.
That's why I keep thinking OpenGradient's biggest competitor isn't another AI model.
It's distrust.
Distrust is what makes people rewrite prompts before sending them. It's what keeps sensitive workflows off AI tools entirely. It's the reason some teams still hesitate even when productivity gains are obvious.
The model race gets most of the attention because it's easy to measure.
Trust is harder.
And I'm not convinced the industry has figured out how to compete on that yet...

What's the biggest barrier to AI adoption? 🤔

#opg $OPG @OpenGradient