June 29, Steve Deng, Chief AI scientist of MATRIX AI Network and Associate Professor at Tsinghua University joined a member of Dorian S. Nakamoto’s original development team on a panel at the “ChainDD Global Roadshow Forum — Beijing.”
Martii Malmi, a core developer who worked on the original Bitcoin project shared an inside glimpse into the mythic figure of “Dorian S. Nakamoto,” whose history has been cloaked in mystery. Nearly one hundred top experts, scholars and tech pioneers in the blockchain field spoke about development trends and emerging applications.
Professor Deng recalled how he first encountered Bitcoin mining back in 2009 through his work and research with GPUs. At the time he hadn’t fully grasped the decisive significance of trust for cryptocurrencies, but came to deeply appreciate how the decentralized, distributed ledger technology underlying Bitcoin was foundational to later abstracted notions of blockchain. Today, the innovation that started with Bitcoin has already expanded beyond a digital currency, to a range of applications in finance, trade, credit reporting, internet of things, the sharing economy and more.
Deng discussed how the superior security, transparency, and efficiency of blockchain, in combination with the use of smart contracts will greatly reduce the burden and repetition of manual recording and auditing, reducing transaction costs in a number of industries. Blockchain will be especially valuable in establishing networks that have enhanced trust, oversight, data integrity, interoperability for collaboration.
Regarding the current state of integrating AI and blockchain, Professor Deng expressed that the technologies share a related fundamental nature. To begin with, blockchain preserves the integrity of data and having high-quality information is critical for useful AI. Furthermore AI requires massive processing power to train and run algorithms, and the mining, consensus, and validation mechanisms draw on a huge pool of processing power. Furthermore, AI is a powerful diagnostic and self-optimizing tool for blockchain. For example MATRIX is using AI for auto-coding smart contracts, security audits and parameter selection. Finally, true intelligence is defined as the ability to adapt to an environment. By combining blockchain’s ability to compile and organize data in space and time with AI’s ability to foresee the future developments from historic data, MATRIX seeks to create a AI-powered blockchain advancing the next evolution of intelligence.
Professor Deng shared that currently artificial intelligence is largely used in centralized systems, which presents several problems. 1) When data processing and data are centrally controlled, there is lowered control of privacy 2) The cost of centralized data storage is constantly rising and yet we’re still not able to meet the global demand for processing power.
Decentralized AI is currently coming of age, with “federated learning” gaining wider acceptance, and arenas where decentralized AI has demonstrated functional advantages. There are two broad areas that are well-suited for decentralized AI computing. 1) Machine learning applications where there is a high imperative to preserve privacy, such as health care. 2) A need for services customized according to individual preferences.
Professor Deng shared that MATRIX is implementing blockchain as a source of processing power for decentralized AI research, and has dedicated substantial R&D efforts to integration of the technologies for enhanced security and performance. At the same time, this technology pairing can be a highly effective reward and punishment mechanism within the ecosystem, using AI to discover positive contributions and bad actors, and blockchain to record this history of conduct on the platform.
Currently the main solution for scaling the volume of transaction data that blockchain can process in a timeframe that meets real world demands is to add the number of processing units in the system. This approach is tantamount to adding additional horses to a chariot. Professor Deng team at MATRIX is progressively moving toward General Artificial Intelligence, representing a nonlinear breakthrough from the current approach, on the order of evolving from a chariot to a spacecraft.
Professor Deng shared that in addition to their current projects, MATRIX is currently working on further innovative applications of AI-powered blockchain for digital media, software/hardware in, asset ownership rights, public service foundation and more. Projects in the pipeline include blockchain infrastructure for the Xiongan FTZ, a media advertising platform, and more; with ambitions to become the world’s largest public blockchain.