Distributed Ledger Technology (DLT) has been through several iterations over the years and currently exists in several forms. The most popular of which, blockchain, was made famous by the cryptocurrency Bitcoin as it grew in popularity over the past decade. Some other DLT concepts which support cryptocurrencies include Nano's block directed acyclic graph (blockDAG) and IOTA's transaction-based directed acyclic graph (TDAG).
However, whichever form of DLT is integrated into modern systems, developers often find they are coming up against the same problems - lack of decentralization, slow transactions speeds, or inconveniently large nodes.
Many popular proof-of-work (PoW) blockchains that host cryptocurrencies like Ethereum grow so large over time that it becomes increasingly difficult for them to be integrated into new systems. This is because the blockchain is programmed to run at full capacity from the moment it's launched, often resulting in empty blocks being created that take up unnecessary space.
Contributing to the Network for A Reward
Another key issue facing blockchains that use a proof-of-stake (PoS) consensus method is the inability to correctly and efficiently select block producers. Unlike PoW, which relies on miners working for rewards, PoS networks select 'block producers' to contribute to the network for a reward. However, once block producers have been selected, they often lack further incentive to contribute to the network, resulting in slow speeds and low usage rates.
Over the past few years, artificial intelligence has exploded in popularity, with several companies like Numerai and Peculium combining AI with blockchain concepts to improve trading accuracy. Other companies like Singularity.NET and DeepBrain Chain have developed open-source, blockchain-based services to facilitate the advancement of AI technology. However, few companies have investigated utilizing AI specifically to resolve the issues affecting how a blockchain operates.
Achieving Blockchain 3.0 - Removing the human element
One way to resolve these issues has been put forward by a new DLT tech startup based in Switzerland that hopes to upgrade blockchain to the next level through AI-integration.
Velas AG is a company founded by Alex Alexandrov, the CEO of long-running crypto payments platform Coinpayments. Alexandrov believes we can overcome the problems found in modern blockchains by removing the human element and automating the network's management tasks with artificial 'intuition,' rather than intelligence. His company has developed the first AI-enhanced distributed proof-of-stake (AI-DPoS) blockchain network, VELAS (Virtual Expanding Learning Autonomous System).
The idea is that 'swarms' of artificial intuitions constantly evaluate every potential block producer on the network, running all necessary data through an algorithm and making the best possible decision for the network at any given moment. As conditions change, the network is slowed down or sped up as needed, and block producers are elected and re-elected to fit the requirements.
"VELAS elects block producers based on performance dependability and amount of estate while also optimizing network parameters such as block production speed, allowing the network to automatically slow down or accelerate, making it capable of processing up to 30,000 transactions per second," said Alexandrov, speaking at the Malta AI and Blockchain Summit 2019.
A move away from traditional AI
Traditional artificial intelligence systems like those developed by Google and Apple work by collecting massive amounts of data and formulating responses based on the information at hand. Some more advanced AI systems integrated with machine learning capabilities will grow and develop their own systems over time in order to provide more accurate responses.
According to Alexandrov, artificial intuition utilizes a more biological approach, similar to the way in which single-cell organisms know how to behave without the need for any input or direction. Referring to a Harvard study on the evolution of bacteria that reveals how cells utilize genetic mutation for survival, Alexandrov highlights the nature of intuition:
"Single-cell organisms are born, and they just kind of know how to do stuff, and that is where the intuition comes in. So we're trying to mimic nature in that sense. The DNA swarms of artificial intuitions that we've created mimic the exact same protocols."
The project not only promises to streamline how blockchains operate but also provide a wealth of valuable information on why specific actions were chosen. By analyzing the output from the artificial intuitions decision-making process, human beings may well be able to evaluate and evolve our own way of thinking. AI and blockchain remain relatively nascent concepts in the world of financial technology, but together, they could represent the future of the sector.
“The crypto space has changed the game in a lot of ways, but it still has a number of problems. AI is the answer to many of these existing issues,” concluded Alexandrov.