April 9, 2018

Blockchain Starting To Feel Its Way into the Artificial Intelligence Ecosystem

James Kobielus


Blockchain is a technology that everybody seems to think will revolutionize the global economy. If nothing else, the cryptocurrency boom has produced a flood of VC money that’s trying to cash in on every potential application for this distributed hyperledger technology.

It’s no surprise that the artificial intelligence (AI) community is also trying to board the blockchain train. What follows are the chief areas where the AI community believes blockchain can provide infrastructural value:

  • Blockchain as an AI compute-brokering backbone: AI developers need the ability to discover, access, and consume distributed computing resources when preparing, modeling, training, and deploying their applications. The Cortex blockchain allows users to submit bids, in the form of AI “smart contracts,” for running AI algorithms in a distributed, trusted on-demand neural-net grid. The Deep Brain Chain uses a blockchain-based cryptocurrency mining environment as a private compute-brokering platform for communities of AI developers, with the currency’s token (a “DeepBrain Coin”) used to bid for compute resources and being earn you serve those resources to others.
  • Blockchain as a decentralized AI training-data exchange: Training data is the lifeblood of AI applications. Ocean Protocol is using blockchain to build a decentralized exchange for AI training data across all industries. In a more device-specialized context, IoTeXhas built a blockchain for sharing IoT data, while DX Network’s blockchain aggregates structured data about companies, investors, and industry news.
  • Blockchain as an AI middleware bus: Distributed AI microservices need to communicate and possibly also share state and persist data. The blockchain implemented by SingularityNETsupports a global networking protocol for communication among distributed AI algorithms and models, and they plan to support cross-domain AI data access, curation, and publishing scenarios.
  • Blockchain as an AI audit log: Tracking, discovery, compliance, and transparency are big issues for AI-based components. Botchain has defined a hyperledger maintaining for tracking the identities and actions of AI-based entities. Within this emerging open-source environment, every AI component can write regular hash functions of their activity to a blockchain that is immutable and introspectible.
  • Blockchain as an AI data lake: Many AI professionals are eyeing blockchain as the hyperledger storage foundation of future data lakes, though we don’t see much adoption yet in this regard. In a similar vein, many are exploring blockchain’s potential value in safeguarding consumer data privacy as a counterweight to AI’s ravenous appetite for such data.


Clearly, none of these is a mature, built-out, widely adopted AI backbone, and many are highly speculative. To the extent that any of these initiatives gains a foothold in production AI environments, it will probably be to support enterprises’ own heterogeneous modeling, training, and deployment pipelines for machine learning and deep learning.

What all of these blockchain initiatives must contend with is the fact that this technology is coming late to game. It’s still not clear what actual value blockchain would add to today’s AI-focused and increasingly mature data lakes, middleware, data exchanges, audit logging, and compute brokering infrastructures.

Beyond the issue of reinventing the wheel, another thing that an AI blockchain has going against it is its orientation toward mesh networks where there is no centralized control. Though there are survivability benefits to be gained by putting AI infrastructure on such a radically decentralized backbone, hyperledgers run counter to the growing move toward establishing a tight governance cordon. What’s to stop a blockchain-based AI malware bot from using that decentralized fabric to scatter its intelligent footprint everywhere, replicate itself in perpetuity, and thereby elude efforts to stamp it out?

Related to that is the potential collision of an AI blockchain with the personal privacy controls dictated by the General Data Protection Regulation (GDPR). Considering that a hyperledger is an unchangeable data record, AI professionals will find themselves in potential GDPR violation every time they put training data, audit logs, and other pipeline data on blockchains where it is next to impossible to delete permanently.

Blockchain boosts the visibility of core data and other assets throughout distributed value chains. For AI professionals, this fundamental fact may expose them to regulatory, business, and technical risks that could prove unpalatable.

About the author: James Kobielus is SiliconANGLE Wikibon‘s lead analyst for Data Science, Deep Learning, and Application Development.

Related Items:

Developers Will Adopt Sophisticated AI Model Training Tools in 2018

Training Your AI With As Little Manually Labeled Data As Possible

Scrutinizing the Inscrutability of Deep Learning

Share This