back
Why We’re Incubating Poseidon
Story
22 July 2025
AI is Getting Physical. The Stakes Just Changed
Over the past year, AI has gone from hype to history in the making. LLMs and image generators built on scraped internet data have reshaped how we work, create, and learn. But what happens when AI moves off the screen and the internet runs out of training data?
The next frontier of AI is not digital; it is physical. From autonomous cars to warehouse robots and smart prosthetics, tomorrow’s most impactful AI systems will need to perceive, understand, and act in the real world.
And yet, for all the progress in models and compute, one thing remains stubbornly scarce: high-quality, rights-cleared training data from the physical world.
Why Data is the Real IP Frontier
As Meta CTO Andrew Bosworth put it, no amount of media content scraped on the internet can replicate “the intuitive judgment of friction and material deformation when people grab a coffee cup.”
You cannot train a robot to make a bed or navigate an intersection using Reddit threads. You need real-world footage, labeled edge cases, and synthetic augmentations with traceable provenance.
But today, physical data is:
- Painful to collect
- Impossible to license
- Offchain, unaccounted for, and unrewarded
This isn’t just a gap, it’s the ultimate test for what we’ve built. We created Story to make all forms of IP programmable, enforceable, and composable in a world increasingly shaped by AI. From art to algorithms, our protocol was designed to register original work, license it onchain, and route royalties across networks of collaborators. Now, we’re extending that infrastructure to its most powerful and underserved frontier: data.
Introducing Poseidon: The Missing Layer for Real-World AI Data
Poseidon is a new project incubated by Story. It is a full-stack data infrastructure layer designed to solve AI’s long-tail data bottleneck.
Poseidon provides:
- Scalable pipelines to ingest real-world data from sensors, videos, and synthetic tools
- A decentralized coordination layer for labeling, curating, and licensing
- Onchain registration of every contributor in the data value chain, from the person capturing footage, to the annotator, to the model that consumes it
It functions as the data operating system for physical AI. Incentives, provenance, and programmable licensing are built into the foundation.
Why Poseidon is Only Possible on Story
The moment a piece of data is registered on Story, it becomes the parent IP. Every contribution after, whether it’s bounding boxes from a human labeler or synthetic variants from an AI system, is linked as a child, forming a lineage of derivative IP with attribution and revenue flows encoded by default.
Royalties are automated. Attribution is enforced by smart contracts. Every relationship in that chain is captured by Story’s IP Graph. What this creates isn’t just labeled data. It’s licensed, monetizable, and composable data that's AI-ready and IP-cleared.
Imagine an AI agent sourcing datasets to fine-tune a domain-specific model. Instead of scraping the open web, it queries Story, filtering for licenses that allow commercial use. Every file it pulls is traceable, usage is compliant by default, and royalties are triggered automatically.
Companies like Scale AI and Mercor built impressive pipelines. But with Meta’s near-acquisition of Scale, the space now has a vacuum.
Model developers need:
- IP-cleared data with provable ownership
- Diverse, long-tail data to improve model training
- Traceable provenance across synthetic and human-labeled assets
Only Story’s Layer 1 blockchain offers all three natively, elegantly, and on-chain.
Why This Needs to Exist
AI does not need bigger models. It needs better data. The arms race in model size is hitting diminishing returns. The next breakthroughs will come not from more parameters, but from more diverse, high-quality inputs. But right now, there is no way to coordinate supply and demand in the open.
For example, imagine an embodied AI startup needs footage of people loading dishwashers to train a home assistant robot. Who collects that data? How do they get paid? Who owns the IP? What happens if that data is used to train a model deployed in millions of homes?
Poseidon addresses these questions. It creates a marketplace where data producers, whether individuals, fleets, or researchers, can contribute long-tail data, register it as IP, and receive rewards when their contributions power downstream AI systems. For developers, it offers access to high-quality, verified data they can actually use, without worrying about copyright violations. It opens the door to rare, real-world datasets, the kind that will fuel the next generation of robotics, autonomous vehicles, and AI agents operating beyond the screen.
Why Now
The AI industry is advancing rapidly, but its most foundational input, data, remains broken. The platforms scraping the web are running out of content. Physical AI systems are being held back not by model design, but by data scarcity and legal uncertainty.
Poseidon is the first protocol-level attempt to change this reality.
It is a system that makes real-world data liquid and usable. Poseidon only works because it is built on Story, which is the only infrastructure that treats data as IP and allows that IP to be compliant at scale.
Why This Team Gets It
Led by Story Founder and CEO, Seung Yoon “S.Y.” Lee as the project’s founding President, Poseidon co-founders Sandeep Chinchali and Sarick Shah bring deep technical credibility, proven execution, and a shared vision for unlocking data as IP.
Poseidon is launching with strong conviction and momentum backed by a $15M seed round led by a16z crypto.
"AI foundation models have already exhausted the most easily accessible training data. Poseidon's decentralized data layer seeks to establish a new economic foundation for the internet, rewarding creators and suppliers for providing the diverse inputs that next-gen intelligent systems need. We are excited to support Poseidon in its work to solve one of the most critical bottlenecks in AI development."
– Chris Dixon, Founder and Managing Partner, a16z crypto
Sandeep, recently appointed Chief AI Officer at Story, is a Stanford PhD, professor of AI and robotics at UT Austin, and one of the sharpest minds working at the intersection of AI, robotics, and distributed systems. With over 1,500 academic citations, his work bridges distributed systems, machine learning, and edge robotics. He’s spent years thinking about how to incentivize rare, real-world data collection, and now he’s building the infrastructure to make it happen.
Sarick, Poseidon's co-founder and Head of Product, brings a product-first mindset to AI systems. Previously the Lead AI Engineer at Story, he designed tools like natural language interfaces (MCP servers) for IP registration and search. Before that, he built multi-agent fleet management systems, causal inference models for telecom, and predictive AI for logistics. His focus is now on building intuitive, developer-facing infrastructure to power new decentralized data economies.
S.Y. Lee, CEO and co-founder of Story, joins as President of Poseidon. He brings company-building experience across multiple ventures, along with the strategic leadership to scale the Poseidon ecosystem and guide its market adoption. Seung Yoon “SY” Lee served as Global Strategy Officer at Kakao Entertainment, a leading Korean entertainment company, after selling his first venture, Radish – a mobile serialized fiction app with millions of downloads – for $440M in 2021. A serial entrepreneur and venture partner at Hashed, he has been recognized as an Asia Society’s Asia 21 Young Leader, inaugural member of Forbes 30 under 30 Asia, a Trilateral Commission David Rockefeller Fellow, and the first Asian President of the Oxford Union.
What’s Next
Today, Poseidon goes live. We believe it will become the coordination layer for the next generation of AI systems, not only in crypto, but across robotics, healthcare, mobility, and more.
If you are building physical AI systems, synthetic training data pipelines, or decentralized data networks, we hope you will build with us.
This is the beginning of something big.
Poseidon is here and ready to unlock the next trillion-dollar dataset.
