In a landmark week for artificial intelligence infrastructure, two pioneering companies announced staggering funding rounds that signal a strategic shift toward solving foundational challenges in the AI era. Runway AI, the video generation leader, secured a massive $315 million Series E investment, while former GitHub CEO Thomas Dohmke's new venture, Entire, raised a record $60 million seed round to manage AI-generated code.
Runway AI's $315 million funding round, confirmed on October 13, 2025, was led by growth equity firm General Atlantic with significant participation from industry giants including Nvidia, AMD Ventures, and Adobe Ventures. This investment catapults Runway's valuation to $5.3 billion, nearly doubling its previous valuation. The capital infusion marks a strategic pivot for the company as it moves beyond its entertainment roots to tackle what many researchers consider the holy grail of AI: building capable "world models."
These advanced systems aim to construct internal, predictive simulations of reality, enabling AI to reason, plan, and interact with complex environments in ways that transcend today's large language models. According to a company blog post, the fresh capital is earmarked explicitly to "pre-train the next generation of world models and bring them to new products and industries." This represents a fundamental evolution for Runway, best known for its Gen-4 series of physics-aware AI video generation tools used widely in media and advertising.
The funding follows closely on the heels of Runway's December 2024 release of its first world model prototype, indicating investor confidence in this high-stakes technological direction. World models differ fundamentally from generative AI tools that create images or text—they are AI systems designed to learn an internal representation of an environment, whether physical, digital, or conceptual. This internal model allows the AI to simulate potential future states and consequences before taking action.
Runway's credibility in this ambitious pivot is bolstered by the demonstrated success of its core video technology. The recent launch of Gen 4.5, its latest video generation model, has been a technical triumph, with industry benchmarks showing it outperforming video generation offerings from both Google and OpenAI. The company plans to rapidly expand its roughly 140-person team across research, engineering, and go-to-market operations, focusing on deploying world model technology in fields like medicine, climate science, energy, and advanced robotics.
Meanwhile, former GitHub CEO Thomas Dohmke's new venture, Entire, announced on October 24, 2025, that it secured a staggering $60 million seed round led by Felicis Ventures at a $300 million valuation. This represents the largest seed investment ever for a developer tools startup and directly addresses the growing crisis of 'AI slop' overwhelming software projects worldwide.
The core mission of Entire is to rebuild the foundational systems of software production for an AI-native era. Dohmke, who led GitHub through the explosive adoption of its AI coding assistant Copilot, identified a critical gap in the developer toolkit. The traditional manual workflow—using issues, git repositories, pull requests, and deployment pipelines—was never designed to handle the volume, velocity, and opacity of AI-generated code.
Entire's technology stack rests on three interconnected components: a Git-compatible database that acts as a unified source of truth for all AI-produced code; a universal semantic reasoning layer that enables multiple AI agents to work together coherently; and an AI-native user interface designed for agent-to-human collaboration. The company's first tangible product is an open-source tool named Checkpoints, which automatically pairs AI-generated code snippets with the full context that created them, including original prompts and conversation transcripts.
The investor syndicate for Entire includes Madrona, M12 (Microsoft's venture fund), Basis Set, and angel investors like Datadog CEO Olivier Pomel and Yahoo co-founder Jerry Yang. This broad support from both top-tier VCs and seasoned industry operators validates the problem space, especially amid an unprecedented 'agent boom' with companies like Anthropic and OpenAI rapidly releasing more advanced, agentic coding models.
Both funding rounds underscore the intensifying competition in AI infrastructure development. Runway's strategic shift occurs within a fiercely competitive landscape where the pursuit of world models has become a central battleground for top AI labs, including Google DeepMind and Fei-Fei Li's World Labs. Similarly, Entire enters a market desperate for infrastructure to harness the power of AI coding agents effectively.
The potential impact of these developments is profound. Successful world models could accelerate drug discovery in medicine, model complex climate systems with unprecedented granularity, and transform robotics training. Meanwhile, effective management of AI-generated code could enable a future where AI agents safely handle routine coding tasks, allowing human engineers to focus on higher-level architecture and creative problem-solving.
