Twitxy Saved

Igor Babuschkin

Verified account River AI

@ibab

CEO & Co-Founder @river_ai_inc. Previously @xAI, Research & Engineering
Palo Alto, CA babushk.in Joined February 2020
902 Following 112K Followers 1.2K Posts

Latest posts 5

Pinned post

Igor Babuschkin Verified account @ibab
We've raised $1.1B to build AI that is owned and shaped by each of us. Check out the article published by the The New York Times that explains River AI's mission and where we're going next.

Our first product is the River API which allows anyone to build custom agents and LLMs based on open weight models: river.ai/api

Congrats to the team and thank you to all of our supporters. Stay tuned for more updates soon.
River AI Verified account @river_ai_inc Aug 11
Today, we're sharing that River AI has raised $1.1 billion, led by @generalcatalyst and @amppublic with strategic investment from @nvidia and @AMD. Additional investors include @ycombinator and @Temasek.

We imagine a future where your AI works entirely for you and deeply aligns with your values. Compared to the corporate chatbots of today, it will feel completely different: responsible, curious, and truly yours. Most importantly, you'll own everything: the hardware it runs on, the data it learns from, and the intelligence itself.

Our training API is already live and powering incredible work today. This new funding accelerates our push to bring the full personal AI stack to life.

Read River AI CEO Igor Babuschkin’s (@ibab) conversation with the New York Times to learn more: nytimes.com/2026/08/11/tec…
View quoted post on X
132 replies 99 reposts 1.6K likes 394K views View on X
Igor Babuschkin Verified account @ibab
By 2028 most of us will rely on powerful AI agents to make our decisions.

After the unprecedented success of coding agents this year, forecasting and decision making are the new frontier. Similar to coding, forecasting is amenable to large-scale reinforcement learning with verifiable rewards: Agents are tested on historical data, predicting past market data and events while only able to access prior events. RL can then teach agents to predict at a superhuman level.

The use of such tools will be irresistible to CEOs, politicians and anyone else who carries responsibility and has the money to afford them. If the agents are demonstrably smarter than you, can you really afford not to use them? These users will be willing to pay several $10,000 per month or more to rely on the agents.

At @river_ai_inc we are working on building open-source forecasting RL recipes that run on our API platform, accessible to anyone. Any company and organization in the world should have the ability to own and improve their own decision-making agents!
41 replies 16 reposts 223 likes 19K views View on X
Igor Babuschkin Verified account @ibab
Jensen leaving Nvidia is a sad day for anyone that cares about open models and making AI cheap, abundant and work for its users.

We need to put in even more effort to ensure we’ll live in a future where you can get things done without asking a big AI company for permission.

Thank you Jensen for all the work you’ve done to foster innovation and entrepreneurship in AI.
22 replies 31 reposts 474 likes 19K views View on X
Igor Babuschkin Verified account @ibab
Define your loss function on your laptop and execute it remotely on hundreds of GPUs
Daniil Ostashev Verified account @vehsatso Oct 30
We just shipped something pretty unique among post-training API providers: end-to-end custom losses.

The objective you define locally gets traced into a graph, compiled and executed efficiently on our GPUs.

Read more about the engineering behind it in our latest blog post.
View quoted post on X
2 replies 3 reposts 86 likes 9.7K views View on X