📍 General Intuition raised $320 million at a $2.3 billion valuation to train AI agents on action-labeled gameplay data. Agility Robotics is moving toward the public markets at a $2.5 billion pre-money equity value. FieldAI has crossed $100 million in revenue and customer contracts. The next AI race is not about who writes the best answer. It is about who teaches machines to act in the real world. ───── Hey Builders! What happens when AI leaves the chat window and starts learning how to move? That is the question underneath this week’s most interesting capital moves. Not whether a model can write, code, summarize, or reason through a prompt. Whether it can understand space, timing, cause and effect, and what to do next when the environment changes. That means knowing what a robot should do in a warehouse aisle. It means understanding what an autonomous agent should do in a simulated factory. It means teaching machines the difference between seeing motion and knowing which action caused it. That is why General Intuition’s new round matters. The company says it raised $320 million at a $2.3 billion valuation to build action foundation models trained on gameplay data from Medal, the video-sharing platform founded by Pim de Witte. TechCrunch reported that the round was led by Khosla Ventures, with General Catalyst, Jeff Bezos, Eric Schmidt, Nico Rosberg, and researchers from Google DeepMind and MIT participating. The investor logic is not only about gaming. It is about a data bottleneck. Large language models were trained on the open web. Robots do not have an equivalent internet-scale corpus of reliable physical action. General Intuition is betting that millions of hours of gameplay, paired with exact records of button presses and timing, can become a cheaper path to models that learn movement, causality, and environment dynamics. The same week brought more evidence that the market is taking physical AI seriously. Agility Robotics announced a SPAC merger that values the company at $2.5 billion pre-money and is expected to deliver more than $620 million in gross proceeds. FieldAI told Business Insider it has grown past $100 million in revenue and customer contracts from 30 customers. Crunchbase reported that robotics startups have already raised $18.8 billion globally in 2026, surpassing the $15 billion raised across all of 2025. For founders, this is not a science-fiction story. It is a market-structure story. The next AI frontier is being shaped by proprietary data, simulation, robotics, industrial customers, embodied systems, and public-market readiness. If software ate workflows, physical AI is now coming for the operating floor. ───── 1️⃣ $320 Million Turns Gameplay Into a Physical AI Training Asset General Intuition said it raised $320 million at a $2.3 billion valuation, bringing its disclosed funding to $454 million, per TechCrunch. The company was spun out of Medal, whose platform lets gamers upload and share clips. The strategic asset is not only the video. It is the action labels embedded in gameplay, including what players pressed and when they pressed it. That detail is the whole thesis. Video can show what happened. Action-labeled video can help a model understand what caused it to happen. General Intuition is using that data to train models for spatial-temporal reasoning, simulation, robotics, and agentic systems that can respond to changing environments. The company plans to use much of the round to scale compute capacity and pre-train the next version of its model, while opening its API more broadly by the end of the summer. TechCrunch reported that General Intuition has a deal with CoreWeave and already has customers in gaming, simulation, and robotics. 💡 For Founders If your AI company depends on proprietary data, make the data advantage legible. Investors are paying for data that competitors cannot easily scrape, license, or reproduce. The strongest pitch is not that you have a model. It is that you have a learning loop no one else can match. ───── 2️⃣ $2.5 Billion Moves Humanoid Robotics Toward the Public Markets Agility Robotics announced on June 24 that it will go public through a merger with Churchill Capital Corp XI. The transaction values Agility at a $2.5 billion pre-money equity value and is expected to provide more than $620 million in gross proceeds, including $420 million from Churchill XI’s trust account assuming no redemptions and approximately $200 million from a common-stock PIPE led by Foxconn. The company says Digit v5 is designed as an AI-enabled, cooperatively safe humanoid robot built for scaled deployment. It has secured more than $300 million of multi-year contracted Digit v5 orders to date, and the proceeds are intended to support existing customer orders, expanded deployments, production scaling, and continued investment in Agility’s integrated platform. This matters because robotics is moving from venture-backed ambition to public-market test. A public Agility would give investors a clearer way to price humanoid robotics, while forcing the company to prove production, customer adoption, safety, margins, and deployment discipline in public. 💡 For Founders Physical products get judged differently than software demos. If your company touches hardware, robotics, logistics, manufacturing, or field operations, build proof around deployments, contracted demand, unit economics, safety, and serviceability. The market will not underwrite vague autonomy forever. ───── 3️⃣ $100 Million Shows Real-World Robot Software Is Finding Buyers FieldAI told Business Insider it has grown to more than $100 million in revenue and customer contracts from 30 customers across Europe, Asia, and the United States. The company builds software that helps robots operate in difficult environments such as construction sites, mines, factories, data centers, defense settings, and energy operations. The commercial point is important. Many robotics companies are still trapped between pilot promise and real deployment because they need real-world data to improve models, but they need useful models to collect real-world data. FieldAI says its Field Foundation Models use physics and probability to make risk-aware decisions in uncertain environments, reducing dependence on task-specific training data. Business Insider also reported that FieldAI raised $405 million last year from investors including Bezos Expeditions, Emerson Collective, Khosla Ventures, NVentures, BHP Ventures, and Intel Capital at a $2 billion valuation. FieldAI’s earlier announcement confirmed the $405 million raise and named many of those investors. 💡 For Founders The deployment loop is the moat. Get into environments where your product performs real work, collects better data, and improves from use. In physical AI, field data is not an implementation artifact. It is the compounding asset. ───── 4️⃣ $200 Million Backs AI Research Automation Outside the Big Labs Mirendil raised a $200 million seed round at a $1 billion valuation, per Tech Funding News and Wall Street Journal reporting, with Andreessen Horowitz and Kleiner Perkins leading and NVIDIA participating. A16z confirmed it is leading Mirendil’s seed round and described the company as building a lab-grade research platform for engineers and researchers outside the largest AI labs. The company’s aim is different from the robotics story, but the market logic is related. The frontier AI labs have internal systems that help researchers run experiments, improve models, manage compute, compare checkpoints, and automate pieces of the research cycle. Mirendil is trying to productize that capability for external organizations. That is a high-risk bet. Mirendil has not yet proved product adoption at scale. But the size of the seed round says investors believe AI research itself is becoming a platform market. If AI systems can help more scientists, engineers, hospitals, universities, and companies run model work directly, the AI ecosystem becomes less dependent on a handful of centralized labs. 💡 For Founders Watch where automation turns inward. The first AI wave automated customer support, content, code, and back-office tasks. The next wave is automating the work of building AI itself. That changes who can compete, who needs talent, and where infrastructure spend concentrates. ───── 5️⃣ $18.8 Billion Makes Robotics a Record-Breaking Capital Category Crunchbase reported on June 22 that robotics startups have already raised $18.8 billion globally in 2026. That exceeds the $15 billion raised across all of 2025 and the $14.1 billion raised during the prior peak venture year of 2021. The category is not one thing. It includes autonomous vessels, humanoids, industrial automation, robot intelligence, defense systems, warehouse systems, and AI infrastructure for real-world operation. Crunchbase pointed to large rounds including Saronic’s $1.75 billion Series D, Neura Robotics’ up-to-$1.4 billion Series C, Skild AI’s $1.4 billion round, and Apptronik’s $520 million Series A extension. The common thread is embodiment. Investors are no longer treating robotics only as expensive hardware. They are underwriting physical AI as a new stack that combines models, sensors, simulation, actuation, industrial workflows, and deployment data. That is a different kind of company than a chatbot wrapper, and it requires a different founder discipline. 💡 For Founders If you are building in robotics, industrial AI, edge systems, logistics, construction, defense, healthcare operations, or data-center infrastructure, do not pitch the category as futuristic. Pitch the specific work your system performs, the budget it replaces or expands, and the evidence that the machine gets better in the field. ───── 🔧 Three moves to make this week 1️⃣ Define your proprietary data loop Write down what data your product collects, why that data improves performance, and why competitors cannot easily get the same signal. If the loop is weak, fix the workflow before you raise. 2️⃣ Price deployment, not demos Physical AI buyers care about uptime, safety, maintenance, support, integration, and field performance. Put those requirements into your commercial model early instead of treating them as post-sale friction. 3️⃣ Map the operating floor Choose one real environment where your product could produce measurable value. That may be a warehouse, clinic, factory, jobsite, data center, port, lab, or public agency. The next AI market rewards builders who understand the room, not just the model. ───── 💬 Quote of the Day "We view this as just the next stage of future pre-training." — Pim de Witte, General Intuition founder, to TechCrunch ───── 🏁 Stay Ahead With OHUBNext This brief is part of what OHUBNext members get every day. For $5.99/month — or $59/year — you get the full daily brief, access to OHUB's Library of Opportunity, self-paced certificates in High-Growth Company Building and Tech Ecosystem Investing, career accelerator tools, and live monthly labs with the OHUB team. The annual plan includes exclusive masterclasses, early course access, and wealth tools built for builders who are serious about owning what comes next. Institutional-grade. Fraction of the cost. 🚀 Join at opportunityhub.co/next ───── 🎬 Closing Thought The AI market is getting physical. That does not mean software becomes less important. It means the winning software has to survive contact with environments that are messy, expensive, regulated, hazardous, and full of edge cases. The frontier is shifting from prediction to action, and action requires a stronger relationship between data, deployment, and accountability. General Intuition’s gameplay-data bet, Agility’s public-market path, FieldAI’s customer-contract milestone, Mirendil’s research-automation round, and Crunchbase’s robotics funding data all point in the same direction. Capital is looking for the next defensible layer after chat and code. It is finding it in systems that help AI learn by doing. For builders, the assignment is practical. Find the environment where your company can learn faster than the market, collect better signals than competitors, and turn deployment into compound advantage. That is where the next durable AI companies will be built. ───── ⚡️ OHUBNext Daily Brief - investments, edge tech, and moves that matter. For 12+ years, OHUB has been building pathways and on-ramps to multi-generational wealth without reliance on pre-existing wealth. Through exposure, skills, entrepreneurship, capital markets, and inclusive ecosystems, we've helped people create new jobs, new companies, and new wealth.