Summary
Highlights
Jeff Cardenas, CEO of Apptronik, discusses his company's mission to create humanoid robots. His personal motivation stems from observing the challenges of aging and the desire to build robots that can perform undesirable tasks, freeing up humans. The discussion introduces Apptronik's robot, Apollo, capable of dexterous manipulation, with initial applications in manufacturing and logistics, and a broader vision for home healthcare.
Apollo is designed to eventually perform a wide range of human tasks, starting with simple manufacturing and logistics. Cardenas outlines a three-stage deployment plan: first, in controlled industrial environments; second, in public commercial settings like retail and healthcare; and finally, in homes for assistive care. He predicts robots will be in homes by 2035, accelerating rapidly.
Apptronik emphasizes owning the core technology, particularly actuators, which are the 'muscles' of the robot. Cardenas explains how modern robots use a network of motors and sensors for dynamic interaction with their environment, moving beyond traditional position-controlled robots. The company has raised $400 million, with Google as a significant investor, focusing on creating the robot's hardware while Google DeepMind develops the AI 'brain'.
A key development in AI is enabling robots to learn, rather than being explicitly programmed for every task. By applying generative AI to robotics, robots can learn from human demonstrations, allowing for rapid adoption of new tasks. This approach makes robots more versatile and flexible, changing the landscape of what robots can achieve and where they can be applied.
Cardenas addresses skepticism about the widespread adoption of AI-powered humanoids, assuring that real-world demonstrations are proving the technology's viability. Apptronik is conducting pilot programs with companies like GXO Logistics and Mercedes. He contends that from a first-principles perspective, humanoid robots can be considerably cheaper than cars, with the potential to be under $50,000, benefiting from matured component technologies from other industries.
Apptronik's manufacturing strategy involves contract manufacturing with Jabil, similar to Apple's model with Foxconn. Cardenas explains that this approach allows them to focus on design and software, learning from past experiences of companies like Tesla, which initially relied on contract manufacturers. While vertical integration might be a long-term goal, prioritizing core competencies and managing cash burn are crucial in the early stages.
Cardenas estimates billions of dollars will be needed for Apptronik's ambitions, acknowledging the capital-intensive nature of the industry. He identifies Tesla and Figure as primary competitors, and highlights China as a significant rival in the global robotics race. China's advantages include massive manufacturing capacity, rapid iteration, and a national robotics strategy with substantial government funding and incentives for adoption, posing a serious challenge to U.S. leadership.
Cardenas emphasizes that robotics will be key to national competitiveness and security. He views the development of humanoid robots as the 'space race of our time,' believing it will fundamentally change the economy by increasing productivity per person. The U.S. needs to lead in this field to shape future policy, solve societal challenges like healthcare and environmental issues, and avoid reliance on other nations for critical technologies and capabilities.