[...The “blockchain - artificial intelligence” pair is capable of providing cardinal changes in all aspects of society, promising the growth of the global economy, despite the fact that they still have a long way to go.]...
The blockchain has many weaknesses, including in terms of security, scalability and efficiency, and artificial intelligence, in turn, suffers from problems related to the possibility of violation of privacy, as well as the lack of trust and the ability to explain the principle of action. The combination of these two technologies seems to be inevitable - they could complement each other to create a fundamentally new generation of digital systems. As shown in the figure, the blockchain will eliminate the need to trust artificial intelligence, provide privacy and explainability, and artificial intelligence will allow building machine learning systems based on the blockchain, which will feature enhanced security, scalability and more effective personalization and management capabilities.
Blockchain for Artificial Intelligence:
Secure Data Exchange and the Artificial Intelligence Market
Decentralized Computing for Artificial Intelligence
Explained Artificial Intelligence
Coordination of interactions between devices that do not trust each other
Artificial Intelligence for Blockchain:
Secure scalable blockchains
Personalization with privacy
Automated Arbitration and Management
Network protection against DDoS attacks based on artificial intelligence
Security and scalability ...
If the attacker does not own most of the mining resources, it is almost impossible to crack the blockchain, however, applications built on the blockchain platform, unfortunately, are not so protected. For example, The DAO, one of the largest crowdfunding platforms that owns Ether cryptocurrency for $ 150 million, was the victim of the theft of $ 50 million. The cracker used a number of mistakes made in the preparation of smart contracts, which allowed to execute repeated transactions and withdraw more money than he invested contracted fund.
Given the rapid progress of machine learning, a blockchain running an intelligent algorithm could very well detect the fact of an attack and automatically turn on defense mechanisms. And if damage is inevitable, artificial intelligence could at least isolate the attacked component from the rest of the blockchain platform. By managing the blockchain, it could improve scalability and increase the reliability of a distributed registry. For example, with a sharp increase in transaction volume, artificial intelligence could automatically increase the speed of creating blocks, which would increase throughput at the price of increased confirmation time.