A New Titan in the Arena: Humans& Secures Historic $480M Seed at $4.8B Valuation

San Francisco, CA — In a move that redefines the economics of early-stage venture capital, Artificial Intelligence startup Humans& (pronounced "Humans And") has emerged from stealth with a staggering $480 million seed funding round. The financing, which values the three-month-old company at $4.8 billion, marks one of the most significant capital injections in Silicon Valley history for a company without a commercial product.

The round was led by SV Angel and company co-founder Georges Harik, with aggressive participation from industry heavyweights including Nvidia, Jeff Bezos, and Google Ventures (GV). This unprecedented valuation underscores a pivotal shift in the AI investment landscape: the consolidation of capital around "super-teams" composed of veteran researchers from dominant labs like Anthropic, Google, and xAI.

The Rise of the "Mega-Seed"

The sheer magnitude of this transaction signals a departure from traditional venture dynamics. Historically, seed rounds—designed to prove a concept—averaged between $1 million and $5 million. Humans&, however, has bypassed the conventional Series A and B stages entirely, securing a war chest that rivals the IPO proceeds of established tech firms.

This capital influx is driven by the astronomical costs associated with training frontier-grade AI models. With Nvidia’s direct involvement, Humans& is expected to secure priority access to next-generation H-series and B-series GPUs, a critical resource that has become the de facto currency of the AI arms race. The backing of Jeff Bezos and Emerson Collective further suggests that the startup is positioning itself not just as a technology provider, but as a fundamental infrastructure layer for the future of digital work.

The "God-Tier" Founding Team

Investors are betting less on a pitch deck and more on the pedigree of the founding team, which represents a convergence of top-tier talent from the industry's fiercest rivals. The roster reads like a "Who's Who" of modern AI research:

• Andi Peng: Formerly of Anthropic, where she was instrumental in the reinforcement learning (RL) and post-training pipelines for the Claude model family (Claude 3.5 through 4.5). Her expertise lies in aligning models with human intent, a core tenet of the new startup's philosophy.
• Georges Harik: A Silicon Valley legend and Google’s seventh employee. Harik was a key architect behind the original AdSense and Gmail systems. His return to an operational role signals a belief that the next wave of AI requires systemic engineering at a global scale.
• Eric Zelikman & Yuchen He: Both former researchers at xAI, Elon Musk’s AI company. They played critical roles in the pre-training and scaling of the Grok chat models, bringing deep experience in handling massive compute clusters and data pipelines.
• Noah Goodman: A dual-threat professor of Psychology and Computer Science at Stanford University, bringing academic rigor to the team’s focus on human-computer interaction (HCI).

Philosophy: Augmentation Over Automation

While competitors like OpenAI and Anthropic race toward "Agentic AI" capable of autonomous operation—effectively replacing human workers—Humans& is carving out a distinct ideological niche. The company’s mission is "human-centric" AI, designed to act as a connective tissue between people rather than a substitute for them.

According to internal memos and founder statements, Humans& aims to build systems that function as an "intelligent group chat" or a collaborative layer. The technology focuses on understanding the nuance of human social dynamics, retaining long-term memory of team interactions, and facilitating decision-making without removing the human from the loop.

This "Humans And" approach challenges the prevailing narrative that AI’s ultimate goal is the obsolescence of human labor. By focusing on collaboration, the startup hopes to bypass the enterprise resistance often met by fully autonomous agents, offering instead a tool that amplifies institutional knowledge and team cohesion.

Market Context: The Talent War Escalates

The valuation of Humans& places it in rarefied air, even among the inflated standards of the current AI boom. To understand the scale of this raise, it is necessary to compare it against other landmark seed rounds in the sector. The table below illustrates how Humans& stacks up against its closest peers and historical anomalies.

Comparative Analysis of Record-Breaking AI Seed Rounds

Startup Name	Founders' Background	Seed Round Size	Valuation	Key Investors
Humans&	Anthropic, xAI, Google, Stanford	$480 Million	$4.8 Billion	Nvidia, Jeff Bezos, SV Angel, GV
Thinking Machines Lab	OpenAI (Mira Murati), Meta, Mistral	$2 Billion	$10 Billion	Andreessen Horowitz, Sequoia
Mistral AI	DeepMind, Meta	$113 Million	$260 Million	Lightspeed, Redpoint
Safe Superintelligence	OpenAI (Ilya Sutskever)	$1 Billion	$5 Billion	Nst, Andreessen Horowitz, Sequoia

Note: The "Seed" designation for these rounds is largely semantic, representing the first external capital raised, despite the sums resembling late-stage growth equity.

Strategic Implications of Nvidia’s Involvement

Nvidia’s participation is particularly noteworthy. As the undisputed kingmaker of the AI revolution, Nvidia’s venture arm has been highly selective, typically backing companies that drive massive demand for compute. By backing Humans&, Nvidia is effectively validating the startup's architectural approach, which likely involves massive-scale model training that will utilize thousands of GPUs.

This partnership grants Humans& a significant logistical advantage. In a market where compute scarcity can bottleneck development, having a direct line to Nvidia hardware ensures that the team can iterate on their "collaboration-first" foundation models without the infrastructure delays plaguing smaller competitors.

The Road Ahead: "Connective Tissue" for the Enterprise

Humans& enters a crowded market but does so with a differentiated product vision. The current ecosystem is saturated with chatbots (ChatGPT, Claude) and coding assistants (GitHub Copilot). However, a platform specifically engineered to "debug" and enhance human collaboration remains largely theoretical.

If Humans& can successfully translate its $480 million war chest into a product that understands context, intent, and team dynamics better than existing tools, it could capture the lucrative enterprise communication market currently dominated by Slack and Microsoft Teams. The bet is that the future of work isn't about AI doing the work for us, but AI helping us work with each other more effectively.

For now, the industry watches with bated breath. Humans& has proved it can raise money like a titan; now it must prove it can build like one.