Responsible AI, Blockchain in Safe and Ethical AI

In 2020 the German-based Pharmact AG has developed a rapid Covid 19 test that delivers results in roughly 20 minutes. Now imagine, if combined with blockchain and AI the speed of diagnosis can be curtailed to under 10 minutes while maintaining the security of personally-identifiable data. After all, the data can be collected and verified on a blockchain ledger, and then the running of AI algorithms affords to create of an integrated platform that enables data from disparate sources to be analyzed. By the end of 2024, about 75 percent of enterprises will operationalize AI, driving a 5x increase in streaming data and analytics infrastructures, according to Gartner. However, this also begs the question of the ethical use of data and privacy solutions.

Centralized Machine Learning is all about creating an algorithm using sample data to identify patterns. All ML works on premise that the more data is fed, the more accurate the forecasting is. The approach to model training is centralized requiring the data to be stored at a central location/ server. The problem here however is whether the user privacy is compromised if not violated. It is for this very reason that General Data Protection Regulation (GDPR) is legislated in many nations.

Federated (Decentralized) learning is a new branch of AI. Here privacy is not compromised as data is not stored in a single location and decentralized computing is utilized. With the help of homomorphic encryption, several clients send their learnings (think of thousands of apps or wearable devices collecting data from thousands of users) to a central server without ever users' personal data. The server and clients keep exchanging data iterations which gets better with each cycle. OpenMinded and Tensor flow federated by Google has started some great work in federated learning

Data Privacy in Blockchain:

Blockchain provides new mechanisms, such as Decentralized Identities (DID) and Zero-Knowledge Proofs, that enable data to be shared in ways that maintain privacy, increased cybersecurity, and more ethical use of personal data.

Decentralized identity (DID) a widely held view among blockchain proponents, refers to self-ownership of personal digital data outside of the databases of the third parties. DIDs reduce the probability of unwanted correlation Unlike an email account, the DID would be owned and stored by a person rather than by an email service provider. Zero-knowledge proofs enable ease of access to identity and other important data while maintaining privacy for individuals. Zero-knowledge proofs use cryptographic algorithms that enable a prover to mathematically demonstrate to a verifier that a statement is correct without revealing any data. A common example is a customer ordering a beer, but the bartender demanding to produce a Valid ID showing whether the consumer is an adult or not. Now if the consumer is to show his Valid ID, he will inevitably also reveal his Blood group, Father's name, etc, information that is not relevant to the bartender and can be misused. Zero-knowledge proofs are powerful tools for maintaining privacy and property control for individuals that may need to provide a bit of personal information but no more than necessary.

Practical Applications of AI intertwined with Blockchain:

Artificial Intelligence (AI) is a broad field that includes machine learning and cognitive computing where computers are programmed to mimic cognitive functions such as learning and problem solving many times faster and more accurately than a human. AI or its subset Computational intelligence, when combined with blockchain systems, can create more robust cryptographic functionality and ciphers thereby making it more difficult for cyber hackers to compromise systems. When blockchain participants have increased control over their data, they have the potential to decide with which parties and for what purposes their data are shared. To collect participant data for use in an AI dataset, participant permissions will need to be obtained. Three practical applications of AI intertwined with Blockchain are:

Smart Grid

The decentralized characteristics of smart blockchains can effectively help smart grids realize the transformation from centralization to distribution. The decentralization of smart blockchain breaks information barriers and realizes secure data sharing among multiple participants.

Internet Vehicles i.e. Tesla

Blockchain can provide trust guarantees, reliable data security, and effective incentive mechanisms, and can develop IoT Vehicles. Blockchain introduces elements, such as cars, people, and service providers. Through its transparency, anonymity, and immutability characteristics, it can ensure mutual trust between different elements, strengthen data information security, and promote data information sharing.

Supply Chain

When integrated with the blockchain, the artificial intelligence platform can discover useful information from point-of-sale sales data, historical purchase data, etc., so that data characteristics can be identified, and predictive analysis can be implemented, including future demand forecasts, and sales model forecasts.

Author: Abhinav Garg - Founders Blocktickets