Applied Intuition launches Dana, an agentic platform that lets users build physical AI workflows through plain-English commands instead of juggling about 20 tools.
Qasar Younis says Applied Intuition has raised about $1bn and never spent it, crediting intentional capital planning and a strong product focus.
Qasar Younis argues physical AI faces safety, real-time and cost constraints that chatbots do not, and predicts physical AI companies will dominate within 25 years.
Applied Intuition wants to put intelligence on a billion machines
Applied Intuition is building the technology layer that puts intelligence on machines across many industries. Physical systems face safety and real-time limits that chatbots do not, and their data is almost all proprietary. The founders argue that a focused provider with a data engine and an agentic platform like Dana is well placed in a very large market.
Takeaways
Physical AI is larger and harder than digital AI. Qasar Younis says machines face safety criticality, hard real-time limits and cost envelopes that chatbots do not. Industrials are about 5% of GDP and automotive 3%. Mining is 1% of workers but 8% of work fatalities.
Labor shortages and risk create demand for machines. The average American farmer is 58 years old and long-haul trucking has record shortages. Younis predicts physical AI companies will dominate in 25 years. Peter Ludwig says AI's real economic impact will come from moving physical systems.
Dana lowers the barrier to building physical AI. Ludwig says models are only about 1% of a full solution. Workflows that once meant switching between 20 tools now run through an agentic interface in plain English. Younis likens it to Cursor and Claude for coding.
General-purpose models fall short for safety-critical machines. Both founders say models from Anthropic or OpenAI are not enough. Ludwig notes almost all physical AI data is proprietary. A data engine feeds failed scenarios back into training, and imitation plus reinforcement learning forms the moat.
It is a technology provider, not a vertical operator. Younis compares Applied Intuition to Nvidia rather than Tesla. It builds and evaluates its own models, sells a development environment for customers' machines, and licenses products. He wants innovative technology and a boring business model.
Customers span many industries beyond automotive. Automotive is a minority, and revenue is evenly split across commercial trucking, defense, construction, mining, agriculture and robotics. Customers are mainly manufacturers, plus mining and port operators. It has over 1,000 engineers and runs L4 trucks in Japan.
Capital raised remains unspent; competitors can coexist. Younis says the company has raised about $1bn and never spent it, crediting capital planning and product focus. He says most rivals can succeed in such big markets, and Waymo sells consumer robotaxis, so it is not a direct competitor.
How the product line grew
Younis and Ludwig, both ex-Google, considered a robotaxi company in the early 2010s and judged the technology and business model not ready.
Tap a name to go there · tap a line to see why they are linked
On this episode 2
Show 1
They founded Applied Intuition in 2017 and started with tools rather than building autonomous vehicles.
Deploying software onto machines became a bottleneck, so the company built an operating system.
It then moved into a vertical autonomy stack with its own models.
It now offers Dana, an agentic platform for developing and deploying physical AI.