Ambient emerges from stealth with $7.2M from a16z CSX to build the world’s first AI-powered proof-of-work blockchain

Ambient, a new Web3 project founded by a Harvard-educated technologist, has emerged from stealth mode promising to completely transform how AI is accessed and deployed around the world.

Backed by $7.2 million in seed funding from Andreessen Horowitz’s Crypto Startup Accelerator (CSX), along with Delphi Digital and Amber Group, Ambient pitches itself as “OpenAI powered by AI Bitcoin.” It’s not just a bold claim—it reflects the team’s ambition to rethink the economics, transparency, and accessibility of AI from the ground up.

AI Meets Proof-of-Work Blockchain: A New Chapter in Decentralized Intelligence

At its core, Ambient is a new kind of blockchain. It blends elements of Solana’s architecture with a fresh take on Proof-of-Work—a system that relies on computational effort to secure a network. But instead of simply processing transactions, Ambient uses that computing power for verified AI inference. That means more accurate, secure AI outputs with a layer of transparency baked in.

The project was built to fix what many see as growing problems in the AI ecosystem: black-box models, lack of privacy, bias, and the risk of manipulation. Unlike closed platforms, Ambient’s infrastructure is open and checkable. The team is working toward something bigger, too—a decentralized, on-chain foundation model that can match or even surpass human-level intelligence, with no central gatekeeper.

“Ambient directly aligns the economics of its network with the generation of useful output to create a focused competitor for closed-source AI unlike anything the market has ever seen,” says co-founder Travis Good, PhD.

The technical stack is no small feat. Ambient supports inference on models with more than 600 billion parameters. Thanks to its unique PoW design, the system cuts verification overhead to 0.1% and slashes training costs by a factor of 10 compared to current platforms. The approach gives miners Bitcoin-style rewards—offering more consistent incentives than Proof-of-Stake networks that rely on transaction fees.

The company’s founding team isn’t new to solving big technical challenges. Ambient was founded by Travis Good, PhD, who previously developed the world’s first mathematically optimal freight railroad movement planner and worked on AI-driven drug discovery and spectroscopy.

Ambient’s CTO, Max Lang, is a serial founder with exits under his belt and past roles at Microsoft and Amazon. In addition to helping Blend Labs land its first major customer, Max played a key role in Cockroach Labs’ 1.0 release and led Vault Health’s pivot to COVID testing, scaling to millions of tests nationwide. The founders’ track record shows they know how to execute at scale.

“Ambient’s talented team has taken a contrarian but powerful approach to tackling the biggest challenges in the field,” says Tommy Shaughnessy, founding partner at Delphi Ventures. “We’re delighted to support them as they uniquely position themselves as a tentpole of the coming agentic economy that seamlessly merges Web3 and Web2.”

Michael Wu, co-founder and CEO of Amber Group, sees Ambient’s infrastructure as a breakthrough for Crypto AI: “Their ability to deliver verified inference at unmatched speeds, while focusing an entire crypto network’s resources on advancing model training, is a true game changer for the future of Crypto AI competitiveness.”

Ambient’s testnet is expected to go live later this year. The launch could mark the start of a new chapter for decentralized AI—where transparency, trust, and scale go hand in hand.