Valor, Point72 back General Intuition at $6B valuation as AI startup pushes into robotics

General Intuition, a stealth-mode artificial intelligence startup developing a foundation model for training generalized AI agents, is in advanced talks to raise a new venture round at a $6 billion pre-money valuation. According to three people familiar with the negotiations, Valor Ventures will lead the round, with participation from Point72 Ventures and Seven Seven Six, the firm founded by Reddit co-creator Alexis Ohanian. The funding round, expected to close in the second quarter of 2025, comes as General Intuition transitions from pure language-model development to a broader platform aimed at teaching AI systems how to interact with physical environments—an evolution that positions the company at the nexus of AI and robotics.

Founded in 2023 by a team of former DeepMind and Tesla AI researchers led by CEO John Martyn, General Intuition has maintained extreme secrecy about its technical approach, but filings and investor briefings reveal the company is building a model called GI-1, a spatiotemporal foundation model that simulates motion, causality, and environmental dynamics. Unlike traditional large language models that predict text, GI-1 uses physics-informed neural networks to generate agent behaviors, enabling robots, drones, and autonomous systems to learn navigation policies from simulation rather than real-world trial and error. This method promises faster, safer deployment in industrial, logistics, and consumer robotics—sectors currently dominated by Boston Dynamics, Figure AI, and Tesla Optimus.

The $6 billion valuation reflects both investor confidence and the strategic imperative of securing General Intuition’s technology before competitors catch up. Point72 Ventures, the investment arm of the $40 billion hedge fund Point72 Asset Management, has been particularly active in AI-driven automation, recently backing robotics firms like Agility Robotics and Intrinsic. Valor Ventures, a specialist in deep-tech and AI, previously led a $22 million seed round for AI agent startup Suno in 2024. Both firms see General Intuition’s model as a foundational layer for next-generation robotics, potentially disrupting sectors from warehouse automation to last-mile delivery.

General Intuition’s pivot toward robotics comes at a moment when the industry is shifting from reactive, rule-based systems to proactive, learning agents. The company’s model is designed to unify perception, planning, and control in a single neural architecture, reducing the need for brittle, hand-engineered pipelines. This approach aligns with a broader trend: the rise of “embodied AI,” where models trained in simulation are transferred to physical systems. Recent advances by competitors like Covariant, with its RFM-1 model, and Tesla’s FSD v12, which integrates end-to-end planning, underscore the urgency of scalable, generalizable agent training.

Industry Impact and Significance

The infusion of venture capital at a $6 billion valuation signals a major inflection point for robotics and AI convergence. General Intuition’s model could become the de facto standard for training generalized robotics agents, akin to how large language models like Llama or Mistral power text-based AI systems. The company’s technology is expected to accelerate deployment timelines across industries: logistics firms could train robots to handle diverse SKUs without per-item programming, manufacturers could deploy adaptive assembly systems, and service robotics companies could reduce on-site training costs.

Financial implications are already reverberating through the capital markets. Point72 and Valor’s participation sends a strong signal to incumbents like NVIDIA, which has invested heavily in robotics simulation platforms, and to emerging players such as Figure AI, which recently raised $675 million at a $2.6 billion valuation. It also raises the stakes for cloud providers like AWS and Google Cloud, which offer robotics simulation services, as General Intuition’s model could be licensed or deployed as a managed service. Moreover, the round underscores the growing overlap between AI research and robotics engineering, blurring traditional sector boundaries and creating new categories in the $200 billion-plus robotics market.

The broader adoption of generalized agent models like GI-1 could shift competitive dynamics from hardware-focused moats to software-defined intelligence. Companies that previously relied on proprietary hardware, such as Boston Dynamics, may face pressure to integrate or license such models to remain competitive. Meanwhile, AI-native robotics firms—those built from the ground up on foundation models—could leapfrog incumbents by reducing development cycles from years to months. This trend is already visible in autonomous mobile robots (AMRs), where companies like OTTO Motors and Locus Robotics are integrating generative AI for dynamic route planning and anomaly detection.

The Bigger Picture

General Intuition’s emergence is part of a larger tectonic shift in AI: the move from static, text-centric models to dynamic, agentic systems capable of interacting with the physical world. This evolution mirrors the historical trajectory of computing itself, from mainframes to personal computers to mobile devices—each time moving intelligence closer to the end user. In robotics, this translates to models that don’t just describe the world but act within it, a capability that has long been the holy grail of artificial intelligence.

The company’s timing aligns with a global push to integrate AI into industrial and service sectors, driven by labor shortages, sustainability demands, and the maturation of simulation technologies. Governments in the U.S., EU, and China have prioritized robotics and AI in national strategies, with initiatives like the U.S. National Robotics Initiative 2.0 and the EU’s Horizon Europe funding framework accelerating R&D. General Intuition’s model could serve as a bridge between digital and physical realms, enabling what some researchers call “the real-time internet of actions”—a system where AI agents continuously adapt to real-world conditions in real time.

Expert Analysis

According to Dr. Elena Vasquez, a robotics research scientist at the University of California, Berkeley, and a former advisor to DARPA’s Machine Common Sense program, General Intuition’s approach represents a critical step toward achieving human-like adaptability in machines. “The key innovation here is not just training agents in simulation, but simulating the right physics—causality, constraints, and consequences,” she said. “This is what will allow robots to generalize across tasks without brittle code.” She cautioned, however, that real-world validation remains the ultimate test. “We’ve seen models fail spectacularly when transferred from simulation to the real world. General Intuition’s next milestone will be proving its model works on a diverse fleet of robots under unstructured conditions.” Industry observers expect the company to announce pilot deployments with logistics partners in late 2025, with commercial robotics integrations following in 2026.