Figure AI appears to have aspirations that reach far beyond the conventional goal of simply developing robotic systems. The company’s ambitions, as articulated by its founder and CEO Brett Adcock, hint at an era when robotics transcends industrial utility and begins to merge with visionary concepts more often found within the pages of speculative science fiction.
During a thought-provoking conversation at the annual Dreamforce conference — an event hosted by Salesforce and attended by leaders in technology and innovation — Adcock engaged in an in-depth discussion with Salesforce’s co-founder and CEO, Marc Benioff. It was here, on a Wednesday afternoon filled with discourse on the future of intelligent machines, that Adcock unveiled a vision that captured imaginations. He described a world in which robots no longer exist merely as tools created by humans but evolve into autonomous entities with the potential to replicate, learn, and collaborate, creating something akin to a ‘new species’ altogether.
Responding to Benioff’s reflections on the parallels between current technological developments and the science fiction universe envisioned in Dennis E. Taylor’s 2016 novel *We Are Legion (We Are Bob)*, Adcock fervently stated, “We’re building a new species here.” His statement, while lofty, encapsulates his belief that humanity stands at the threshold of a transformation where artificial beings might someday share qualities of reproduction and collective intelligence. He elaborated that the possibility of machines capable of building and improving upon themselves could arrive far sooner than skeptics expect.
Adcock drew particular inspiration from the theoretical construct of self-replicating ‘von Neumann probes,’ conceptualized by the mathematician and physicist John von Neumann decades ago. These hypothetical machines, which possess the capacity to construct copies of themselves from raw materials encountered in space, form the foundation of Adcock’s futuristic outlook. He painted a detailed picture of an era in which robotic explorers could traverse distant galaxies, extracting resources such as methane from alien planets, assembling replicas of themselves, and effectively participating in the large-scale expansion — even colonization — of the cosmos. This intriguing synthesis of robotics, artificial intelligence, and space exploration, he suggested, might not remain confined to distant speculation but could very well take shape within our own lifetime.
Adcock concluded with quiet conviction, stating that he truly believes these advancements will manifest within the current generation’s horizon.
OpenAI, the globally recognized artificial intelligence research company, appears to share in much of Adcock’s optimism regarding what the integration of AI and robotics could achieve. The company has invested five million dollars in Figure AI, signaling strong confidence in its trajectory. Sam Altman, OpenAI’s CEO, has expressed complementary perspectives, emphasizing in a May interview with Bloomberg that the age of everyday coexistence with humanoid robots is imminent. Altman predicted with excitement that soon people will casually encounter several robots performing practical tasks in everyday life — an experience that, while commonplace, will still evoke the uncanny sensation of living inside a science fiction narrative. He observed that this transition, where machines take on the responsibilities once belonging exclusively to humans, will provoke profound emotional and societal reactions.
However, Figure AI’s pursuit of this cutting-edge vision unfolds amid an increasingly competitive and turbulent landscape, particularly involving figures as influential as Tesla’s CEO Elon Musk. Musk, whose interests in artificial intelligence have long intersected with those of OpenAI, became embroiled in a series of legal disputes against the organization. In 2024, he filed a lawsuit accusing OpenAI of deceiving him into cofounding the enterprise. The company responded with a counter-suit the same year, asserting that Musk’s actions constituted harassment. Tensions escalated further in September, when Musk — through his company xAI — initiated yet another legal complaint, this time accusing OpenAI of misappropriating proprietary data by recruiting key technical employees, a charge that OpenAI has publicly denied.
Simultaneously, Tesla continues to develop its own humanoid robot, known as Optimus. The first prototype of this human-like machine was demonstrated during Tesla’s ‘AI Day’ presentation in 2022. Musk later suggested that Optimus could become available as a consumer product by late 2025 or 2026, heralding a new phase in Tesla’s integration of artificial intelligence into physical robotics. Meanwhile, Figure AI innovates at a rapid pace; in October, the company revealed its third-generation humanoid model, dubbed Figure 03. This newest iteration was designed primarily to assist with domestic responsibilities such as household chores, though it remains out of reach for everyday consumers at present.
The philosophical question of form — whether robots should resemble humans or not — surfaced toward the end of Benioff and Adcock’s dialogue. Benioff mused aloud about whether the ultimate design of intelligent robots might mirror human beings’ familiar anatomy or perhaps take alternative mechanical shapes, reminiscent of the transforming machines popularized by the *Transformers* film franchise. This reflection provoked a broader conversation across the robotics community: must intelligent machines imitate the human body to achieve efficiency, or could other configurations prove superior in practice?
One notable example that illustrates the diversity of design is found in the innovations of Diligent Robotics, an Austin-based company that approached the issue from a distinctly utilitarian standpoint. Its flagship creation, Moxi, is a wheeled robot developed specifically for healthcare environments. Unlike humanoid models designed to walk on legs, Moxi operates on wheels yet performs essential logistical duties such as transporting lab specimens, delivering medical instruments, and managing routine errands for hospital staff. According to the company’s cofounder, Vivian Chu, Moxi represents what she calls the ‘minimum viable humanoid’ — a form that minimizes unnecessary complexity while maximizing functional capacity.
In a prior interview with *Business Insider* reporter Lloyd Lee, Chu explained that the combination of simple mobility — two arms, a head-like interface, and a stable wheeled base — might prove to be the most practical design framework for the majority of real-world applications. Over time, she predicted, such efficient and adaptable robots will likely dominate professional and domestic environments, illustrating that the evolution of robot design is not solely about mimicking humanity’s appearance, but about capturing its efficiency, adaptability, and capacity to assist.
Ultimately, both Adcock’s sweeping cosmic vision and the grounded pragmatism of innovators like Chu reveal the vast spectrum of imagination driving today’s robotics revolution. Whether humanity witnesses the rise of self-replicating, spacefaring machines or simply the normalization of efficient human-assist robots on Earth, one theme remains clear: the boundary between technology and intelligent life is becoming increasingly permeable, ushering in a future that only recently belonged to the realm of fiction.
Sourse: https://www.businessinsider.com/brett-adcock-figureai-humanoid-robot-new-species-dreamforce-2025-10
