Brett Adcock's latest venture, Hark, has secured more than $700 million in Series A funding at a $6 billion valuation, according to a report by Bloomberg. The round closed approximately two months after the company emerged from stealth, placing Hark among the most highly valued AI hardware startups before it has shipped a single product.
Parkway Venture Capital led the round, with a notable investor list that reads like a who's who of the chip and cloud ecosystem: Nvidia, AMD Ventures, Intel Capital, and Qualcomm Ventures all participated, alongside Salesforce Ventures, Brookfield, ARK Invest, Greycroft, Prime Movers Lab, Align Ventures, and Tamarack Global. The presence of both GPU giants and semiconductor leaders highlights the strategic importance of Hark's ambitious vision.
Adcock's track record
Brett Adcock is no stranger to building high-stakes hardware companies. Before Hark, he co-founded Vettery, a recruiting marketplace that was acquired by Adecco for $100 million; Archer Aviation, an electric aircraft manufacturer that went public via SPAC in 2021; Figure, a humanoid robotics company where he remains CEO; and Cover, a school-security company. He also self-funded Hark with $100 million of his own money in late 2025.
Adcock's experience at Figure and Archer gives him credibility in shipping physical products at scale. Figure's humanoid robot, Figure 02, has been deployed in BMW manufacturing facilities, while Archer's Midnight eVTOL aircraft is progressing toward certification. This hardware expertise is central to Hark's thesis: that integrating silicon and models from day one is the most defensible approach in the rapidly evolving AI hardware landscape.
What Hark is building
Hark describes itself as developing a 'personal AI platform' that pairs in-house foundation models, software, and native hardware with new interfaces. The company has deliberately avoided committing to a single layer of the stack, instead building a vertically integrated system that spans model training, silicon design, and end-user devices.
According to the March 2025 announcement on BusinessWire, Hark intends to release its first multimodal models this summer. These models will be capable of processing text, images, audio, and potentially video, forming the foundation of the company's hardware ecosystem. What remains undisclosed is the actual form factor of the device: whether it will be a wearable, a pocketable AI companion, a desktop hub, or something else entirely.
Hark has also not revealed headcount, target price, launch market, or customer pipeline. The $700 million Series A gives the company significant runway to keep its plans under wraps while it iterates on product development.
Capitalizing on a crowded but failure-prone market
The category Hark is entering is small, capital-intensive, and littered with cautionary tales. Humane's AI Pin, launched in 2024, became the most public failure in the wearable AI space, with poor sales and negative reviews. The Rabbit R1, another attempt at an AI-powered personal assistant device, also struggled to gain traction. Even Apple, with its unmatched hardware distribution and control over the iPhone ecosystem, has spent years trying to define its on-device AI offering, recently introducing Apple Intelligence with mixed reception.
What sets Hark apart from these predecessors is its vertical integration. Humane and Rabbit relied on third-party models and off-the-shelf chips, limiting their ability to optimize the full user experience. By designing its own silicon and training its own models, Hark can fine-tune every layer of the stack for low latency, privacy, and energy efficiency. Nvidia and AMD's presence on the cap table also ensures supply allocation, often a binding constraint for AI hardware startups in 2025 and 2026 as demand for advanced chips outstrips supply.
The broader AI hardware landscape
Hark's fundraising comes at a time when venture capital is pouring into AI infrastructure at unprecedented levels. In 2024 alone, AI startups raised over $100 billion globally, with a significant portion going to hardware companies building custom chips for inference and training. Companies like Groq, Cerebras, and SambaNova have raised billions to challenge Nvidia's dominance, but they focus on data-center AI rather than consumer or edge devices.
Hark is targeting the edge—the personal device that runs AI locally without relying on cloud connectivity. This is an area where privacy advocates see immense potential, as it reduces the need to send personal data to remote servers. It also addresses latency concerns for real-time interactions, such as voice assistants or augmented reality overlays. However, the constraints of battery life, thermal management, and processing power at the edge have stymied many attempts to deliver a compelling consumer AI device.
Adcock's approach is to start with the model and work backward to the hardware. By training a foundation model that is efficient enough to run on low-power silicon, he can design a chip that maximizes performance per watt. This is reminiscent of Apple's approach with the Neural Engine in its A-series and M-series chips, but Hark aims to go further by creating a dedicated AI-first device rather than augmenting an existing product category like the smartphone.
Challenges ahead
Despite the massive funding and strong investor backing, Hark faces significant hurdles. Product-market fit is not guaranteed. The company joins a category where several well-funded, well-credentialed teams have launched, missed, and quietly retrenched. By Hark's own timeline, the first models are weeks away, but the device that turns those models into a business is still further out.
Building a custom chip is a multi-year endeavor that requires specialized engineering talent—talent that is in high demand across the industry. While Hark's valuation and funding give it resources to compete for top engineers, the company must also navigate the complexity of manufacturing with foundries like TSMC, securing wafer allocation, and managing yield issues that plague first-generation silicon.
Another challenge is differentiation. If Hark's device ends up being a smartphone-like wearable or a smart-home hub, it will compete not only with Humane and Rabbit but with established players like Samsung, Google, and Apple. Even with vertical integration, convincing consumers to adopt a new device category is an uphill battle. The history of consumer hardware is filled with innovative products that never achieved mass adoption—Google Glass, smartwatches before the Apple Watch, and countless VR headsets.
Adcock's track record gives investors confidence that he can navigate these challenges. At Archer, he led the company from a concept to a publicly traded company with a certified aircraft. At Figure, he has deployed humanoid robots in real-world manufacturing environments. These experiences have taught him to manage supply chains, raise capital in difficult markets, and pivot when necessary.
What the funding means
The $700 million Series A is one of the largest funding rounds in AI hardware history, surpassed only by a handful of data-center chip startups. It values Hark at $6 billion before generating any revenue, a testament to the market's belief in Adcock's vision and execution capabilities. The participation of Nvidia and AMD is particularly noteworthy, as both companies invest selectively in startups that could either expand their ecosystem or, in the long term, become competitors. For now, Hark's need for compute and supply allocation aligns with Nvidia and AMD's interests, so the relationship is symbiotic.
With the funding secured, Hark can now focus on the difficult work of turning its platform into a real product. The company has the capital to hire top engineers, secure manufacturing capacity, and run extensive testing. The next milestones will be the release of its multimodal models in summer 2025 and the reveal of its hardware form factor, likely later in the year or early 2026.
For the broader tech industry, Hark's rise is a signal that the race for personal AI hardware is far from over. While the first wave of devices stumbled, the lessons learned have informed a new generation of vertically integrated efforts. Whether Hark can succeed where others have failed remains to be seen, but the company has the funding, the talent, and the founder to give it a serious shot.
Source: TNW | Investors-Funding News