👋HumanChain - Data Infrastructure for AI Safety
Vision
Build a safe, trustworthy, and human-centric digital world for the AI era.
Why?
AI safety concerns are staring us in the eyes due to exponentially improving LLMs, Gen AI, and Robot capabilities. Examples:
Unconsented model training
Legal and Regulatory compliance (AI act)
Data Privacy and Control
Deepfakes
Identity Theft
Persona Cloning
Mass Propaganda attacks & Threat to democracy
Misinformation
Loss of Privacy and Autonomy
Financial instability
Increase in Unemployment
Discrimination and Bias
AI driven scams
Excessive robot penetration in the physical world
When?
We are focused on the pre-AGI era problems but also laying the foundation for the post-AGI world.
What?
Infrastructure for AI safety consisting of:
A secure and privacy-preserving identity layer
A data layer controlled by users
A global value exchange layer
A post-quantum-safe security layer
An application layer
Tiered identity system where each tier unlocks different data rights for humans and the scope to add Robo ID in the future. Open, permissionless, interoperable infrastructure of data owned and controlled by people. Through its $HUC utility token, it makes data a truly liquid asset class motivating data providers and consumers to exchange data while maintaining compliance and consent.
The incentivisation of end users via data control benefits and tokens makes them an active participant in the data ecosystem.
Developers can train AI models or build applications on top of the application layer without worrying about Bots vs Humans, KYC, Data Compliance & Quality, and copyright issues, and strike an equilibrium between personalised user experience vs data privacy.
How?
We will take a product-first approach and build a few AI Safety use cases to attract users and enterprises to seed the platform. Once we reach a critical mass of data providers and consumers, we transition to a decentralised data exchange. Post that we will launch HumanChain Infrastructure with defined protocols and standards for anyone to build their AI safety applications and integrate AI Safety SDKs within their Web2 or Web3 projects.
Example applications that can be built or supported on HumanChain:
Zero Knowledge Data Rooms to train AI models with zero privacy loss
Federated fine-tuning of AI models to provide personalised experience
AI Avatars or Persona Likeness Licensing and Authenticity
Bots Vs Humans
Deepfake Detection
LLM Wrappers
Personal AI assistants
Regulatory Compliance and KYC
Content Provenance
Decentralised Data Marketplace
Creative Surge: Artists and NFTs
Review or Reputation Systems
Matchmaking platforms
Sales Prospecting and Marketing
Who?
Anchal Taatya is an IIMA alum and 4x founder with 2 exits. His last startup Sparklehood is a profitable venture at the intersection of community, executive hiring, and sales activation. His FMCG venture Goldenberry achieved $12M ARR in the 1st year. His fintech startup Circles, a credit card benefit exchange platform, is a case study at Harvard Business School.
He has also been the world's largest operator of Worldcoin, running in India, Singapore, and Argentina.
Saurabh Kumar is an IIT Kanpur alum and 2x founder. His previous startup Sparklehood is a profitable venture at the intersection of community, executive hiring, and sales activation. Before that, he built Ziffyhomes, a curated marketplace of serviced apartments, incubated at YC (S18) and valued at $40M.
Over the last 8 years, he has grown 10+ communities from scratch to a size of up to 50K members.
Gaurav Mittal is a 2x founder and seasoned tech leader. He's been researching data privacy in AI for the last 1 year. Previously, he built Betwixt, a mobile game at the intersection of NFTs, generative art, and mental health that raised $1M and is now growing organically.
Gaurav is an Elite gamer (WoW rank #1, Hearthstone Top 50). He is also a student of Austrian economics and a community partner at a micro VC fund.
Roman Yampolskiy is an advisor, computer scientist, and an AI safety global thought leader.
Mark Settle is an advisor and ex-CIO of Okta, the world’s leading identity management company.
Mike Massari is an engineering leader, Director of Platform & Integrations at Chainlink Labs.
Nell Watson is the chairperson of the IEEE AI standards committee and an AI philosopher.
Ram is a senior solutions engineer at Polygon Labs and was leading the Polygon ID project.
Bryan Clark is a senior consultant at Informatica, the world’s largest data management company.
Alberto Duso is Strategy & Product Head at Avaloq Labs, Europe's leading fintech solutions provider.
Read our whitepaper
Write to us if you'd like to contribute
Last updated
