Artificio's thesis is that in order to train general purpose Self-Driving Cars, we must rely on 3 key innovations that no other company is focusing on in tandem:

1) Leveraging Open-Source Vision Language Action (VLA) models as base models for driving reducing the training cost CapEx;
‍
2) Crowd-Sourcing Data from taxi drivers (and partnering vs competing with them) across multiple cities in the world through dash-cam data;

3) Leveraging Out-Of-Distribution driving data from emergent economies where wild and unpredictable behavior serves as proxies for adversarial training & edge-case scenario testing.

These 3 pillars plus our new revenue model will allow us to create lower cost self-driving cars as RoboTaxis that can be deployed in any city in the World starting with LatinAmerica.

Follow our latest news and Progress on LinkedIn