We’ve all marveled at how chatbots grasp the world, but what if they were intricately linked to the actual world? Imagine if the dataset powering the chat interface was the physical reality itself, captured instantly by interpreting data from billions of sensors scattered worldwide. This concept underpins Archetype AI, an ambitious startup that debuts today. Ivan Poupyrev, the co-founder, and CEO, envisions it as “ChatGPT but for physical reality.”
Archetype’s core model, named Newton, is engineered to process information from diverse sensors to offer insights, generate charts, or even provide code snippets to depict real-world occurrences. For Poupyrev, Archetype embodies his enduring belief that the digital realm can deeply engage with the physical domain. Interestingly, my book Hackers played a role in sparking Poupyrev’s fascination with altering the world through technology, a passion kindled when his father received the book during a trip to China. Poupyrev’s proficiency in English was also honed through avid reading and rereading of the book.
Poupyrev’s professional journey post his departure from the Soviet Union, transitioning into a computer scientist, encompassed roles at Sony, Disney, and until recently, Google’s ATAP division. His tenure at Google involved spearheading the Soli project, integrating miniature radar devices into wearable tech to enable gesture-based interactions. Transitioning to Archetype, Poupyrev and a team of four colleagues ventured to explore the potential of leveraging large behavior models to process real-time sensor data from the physical world. This endeavor led to the inception of Archetype, supported by a substantial $13 million seed funding round.
Brandon Barbello, Archetype’s COO and co-founder, emphasizes the complexity of the physical world and the challenges it poses due to the intricacy and pace of events surpassing human perception. By harnessing AI to interpret sensor data, Archetype aims to unravel these complexities, enabling a deeper understanding and resolution of pertinent issues.
Archetype’s AI model, Newton, assimilates data from various sensors, translating it into straightforward descriptions of real-world events. This capability extends to diverse applications, from conversing with a household to interacting with a factory, offering real-time insights akin to a comprehensive digital mirror of the entire facility.
The startup has garnered support from Amazon’s Industrial Innovation Fund, reflecting the potential to optimize logistics operations and enhance delivery efficiency. Moreover, collaborations with entities like Volkswagen and healthcare institutions underscore the broad utility of Newton in diverse domains, from assessing recovery post-surgery to enhancing automotive safety through behavior analysis.
While the prospect of a sensor-driven monitoring system raises concerns about privacy and surveillance, Archetype emphasizes its focus on leveraging benign sensor data, such as radar inputs, to address specific challenges without compromising privacy. Volkswagen’s emphasis on behavior detection rather than identity underscores the ethical considerations embedded in Archetype’s approach.
Archetype’s vision aligns with a broader trend in the robotics domain, where companies explore the synergy of generative AI and sensor technologies. Despite competition, Archetype distinguishes itself through its comprehensive approach, aiming to make sensor data universally functional and valuable across various applications.
As Archetype advances towards its mission of comprehending the physical world, the possibilities and challenges that ensue prompt contemplation of the new frontiers that await once this mission is accomplished.
