I've spent the past decade building AI at DeepMind, OpenAI, and xAI, scaling up deep learning and RL. The takeoff of AI progress happened faster than I expected. Years ago, I started the Alphacode project as my brainchild. Then over the span of a few years, coding models became superhuman! Soon AI will be capable of most of the economically valuable work that people do.
There's a question that's been keeping me up at night: if machines can do the work, what is left for people? We started River AI to answer this question.
We don't think replacement is inevitable. Today's agents already work alongside people. But they don't learn from you, and you don't really control them. However, a new type of AI that learns from you, and that you actually control, can lead somewhere very different. Imagine kicking off an agent on a long task and not having to babysit it: it understands your style, your preferences, and your goals, and does the work as if you had been supervising the whole time. That is the kind of AI we are building at River.
To get there, we believe the stack has to be rebuilt end to end: training, models, the product layer, and new hardware that lets personal AI live close to you. Today's stack was designed for a world where a few companies own the intelligence and rent it out by the token. Personal AI does not fit that shape. It is yours, not rented, and you have real control over it.
The River API is the first piece we are sharing. It sits at the training layer and lets anyone build and adapt their own models. If you have ever wanted to take a strong base model and shape it around the work you actually do, this is for you. The next layer we are building, on top of the API, is personalization and continual learning for agents. That is what turns a strong base model into one that actually knows you.
An AI that knows you deeply, understands your goals, and can be trusted to act in your interest is, by definition, highly aligned. We don't think of alignment and safety as separate problems to solve. They are the natural result of building AI around the person it serves.
Capable agents will be a normal part of everyday life. Less like the assistants you call on today when you need a task done, more like guardian angels: quietly present, on your side, helping with what actually matters to you. They will know you well, and they will be yours, not someone else's.
This kind of company only works with the right team. We are looking for people who can do work at the top of the field, who care about how it lands in the world, and who have the taste to ship products that feel finished. We keep talent density high and the team small on purpose. If that sounds like you, come build with us.
— Igor Babuschkin, Co-Founder & CEO