Genesis AI, backed by Khosla Ventures, has unveiled GENE-26.5, its first foundational model for robotics, alongside a demonstration of robotic hands executing complex tasks.
Genesis AI revealed its initial offering after securing $105 million in seed funding to develop foundational AI systems for robotics applications.
The startup's debut model, GENE-26.5, represents the company's attempt to create generalized AI architecture applicable across robotic systems. Rather than task-specific solutions, foundational models aim to provide a flexible base layer that developers can adapt for various robotic applications.
The demonstration showcased robotic hands performing intricate manipulations, suggesting the model can handle dexterous control tasks. This "full stack" approach indicates Genesis AI is building both the AI models and the hardware integration needed to deploy them effectively.
Foundational models have transformed software development, with large language models like GPT serving as base layers for countless applications. Genesis AI's strategy applies this pattern to robotics, a field where hardware and software have traditionally remained siloed.
The company joins other ventures pursuing similar paths, including Boston Dynamics and Figure AI, which have also announced advances in general-purpose robotic systems. The robotics industry increasingly views foundational models as essential infrastructure for scaling autonomous systems beyond narrow, single-task applications.
Genesis AI's $105 million seed round places the startup among well-funded robotics ventures, though specific details on valuation and investor composition remain limited. The funding reflects investor confidence in foundational AI approaches for robotics, despite the field's history of slow commercialization timelines.
The GENE-26.5 demo serves as proof of concept for the model's capabilities, though typical development cycles suggest broader commercial deployment remains some distance ahead. The startup's next milestones will likely focus on expanding task diversity, improving reliability, and attracting enterprise partners willing to integrate the foundational model into production systems.
Bumble's paying user base is declining as the dating app prepares a significant redesign later this year. The company is betting that overhauling its core swiping model will convert more matches into actual dates.
Nous Research, an AI agent developer, is securing at least $75 million in new funding at a $1.5 billion valuation. Robot leads the round with participation from Union Square Ventures and other investors.
General Fusion completed its Nasdaq debut through a reverse merger, becoming the first publicly traded fusion company. The listing marks a milestone for the Canadian firm pursuing commercial fusion energy.
Richard Sutton, the 2024 Turing Award recipient and pioneer of reinforcement learning, has founded Oak Lab in Toronto to develop AI agents capable of continuous learning from their environments.