OORT, a decentralized blockchain-based cloud computing platform for artificial intelligence, today announced the launch of HumanAIx, a global alliance for decentralized protocols to build better AI infrastructure.

HumanAIx will be an open protocol where partners will connect together, each contributing a critical component – validation, storage, compute and data — to form a decentralized AI infrastructure that will be permissionless, scalable and verifiable. OORT said it will allow its participants to remain in control of their data, manage model creation, training and privacy while receiving fair compensation for their work.

“We believe AI should create more benefits than constraints for humanity,” said Michael Robinson, chairman of the OORT Foundation. “With the launch of HumanAIx, we are fostering a decentralized movement to ensure AI remains human-centric and serves the broader community.”

On the validation side, alliance partners include DaTaSECURE and Unyt, which will provide cybersecurity and trustless mutual accounting for AI contributions. Holo Host will supply decentralized hosting and ICP will provide the main interface for access to AI models, datasets and compute.

To scale the decentralized data network layer, XYO will power the HumanAIx AI training and validation network and AI Kido will contribute to its development with a decentralized data collection protocol. Decentralized cloud storage platform Storj Labs Inc. will provide secure, permissionless storage built atop OORT’s node architecture.

AI compute will be delivered by confidential blockchain decentralized cloud Oasis Labs Inc., graphics processing unit network for distributed training protocol provider Io.Net Inc. and graphics processing unit cloud company Valdi Labs PBC.

Unlike large-scale centralized AI, the promise of decentralized AI promises to offer greater data privacy, security and transparency by distributing data processing across a network of nodes rather than relying on a central server. No single entity can control the network and thus can be breached. The use of encryption and controls delivers increased participation in the process for developers and providers of data.

Currently, data labeling and curation rely on underpaid workers and poor working conditions, where many employed workers don’t receive fair wages for their data. OORT said with HumanAIx, by adding extra controls to data added for training, individuals providing data could have their work monetized fairly when their artwork, words or labeling was used in training. That could help to alleviate potential ethical or human rights issues with the training of AI models.

“At XYO, we are driven by the vision of blending blockchain, data and AI to unlock human potential in meaningful ways,” said XYO co-founder Markus Levin. “HumanAIx offers a unique opportunity to foster collaboration and build systems that prioritize fairness, authenticity and real-world impact while ensuring data integrity.”

Images: Pixabay, OORT