(BlockchainConnect Conference, 2018)
DeepBrain Chain is the first artificial intelligence platform to be driven by blockchain technology. The DeepBrain cloud platform was released in April 2017 and the DeepBrain Chain Token Sale whitepaper was released in August 2017. The DeepBrain Chain platform is described as decentralized, lost-cost, and privacy-protecting (DeepBrain Chain, 2018). DeepBrain Chain is designed to save organizations time and money when training AI models and performing AI processing. DeepBrain Chain is building both an AI platform and a smart economy system.
Problems DBC Aims to Solve
- The training of AI models through neural network calculation is currently very time-consuming and expensive.
- Large user bases and variations in usage across time periods drive up costs and often result in idle computing resources.
- There are issues with the data used in AI applications — due to issues of data privacy and proprietary concerns, sometimes users who purchase the rights to use data cannot access the data in a detailed enough manner to use it for AI training.
How DeepBrain Chain Works
The decentralized nodes in the DeepBrain Chain network consist of large clusters of permanent nodes, independent nodes of idle GPU servers, and idle GPUs owned by individual persons. Transactions across the DeepBrain Chain network use smart contracts based on NEO and the DeepBrain Chain token, referred to as DBC. Miners will be rewarded with DBC when they perform computations.
(DeepBrain Chain, 2018)
When a company wishes to use the DeepBrain Chain platform, they will submit their neural network, calculate the demand to be placed on the DeepBrain Chain network, and submit the required amount of DBC GAS, which is the payment required for using the resources of the network. This client request is turned into a docker image, which is then broadcast to the DeepBrain Chain network. The mining nodes in the network will work to deploy the docker image, and the nodes that successfully deploy will receive DBC as a reward.
(DeepBrain Chain, 2018)
Each mining node consists of 6 core parts: a computing engine, an image management system, a storage management system, an identity service engine, a network management engine, and a database service engine. These 6 parts interconnect to work toward deploying the docker image.
Mining nodes are only rewarded if they successfully deploy images and their compensation is based on their level of success. Nodes that exit in the middle of training will be punished instead of rewarded by having their probability of rewards reduced for a 100-hour time period (DeepBrain Chain, 2018).
For more details about the core parts of a node and DeepBrain Chain’s mining algorithm, see DeepBrain Chain’s whitepaper.
(DeepBrain Chain, 2018)
DeepBrain Chain allows sellers and buyers to interact across a decentralized data trading platform. In this platform, sellers can send their data to anonymous nodes to be trained and then worked on by computing engines. Buyers will be able to access the results of their request without ever having to access the raw data that seller allowed to be trained. This allows the seller to continue to own the data while selling the ability to use it with a much-decreased risk of having the data stolen.
The bottom-layer blockchain of DeepBrain Chain helps govern the platform. This bottom-layer blockchain uses a Matrix platform and high-performance asynchronous architecture, has a multi-chain framework, uses a self-encoded two-layer transport protocol or a two-layer encrypted transport protocol, has a multi-layer network, and uses a consensus mechanism that is a hybrid of delegated proof of stake (DPOS) and proof of importance (POI) (DeepBrain Chain, 2018) — for more technical details, see DeepBrain Chain’s whitepaper. According the whitepaper,
“Processing is done in a multi-layer, process-as-they-come method, via the consensus, verification and storage layers to make the entire network is more elastic” (DeepBrain Chain, 2018).
DeepBrain Chain is designed to be used primarily for AI training, modeling, computing, algorithms, and data. For example, a small organization seeking to implement an image recognition solution might choose to use DeepBrain Chain to train their AI in order to save time and money as compared to purchasing to the components necessary to build their own system.
Another promising use case of DeepBrain Chain involves its data trading platform. A seller who wants to sell usage of the data they own could use the data trading platform to allow buyers to use their data for AI training applications while the seller is confident in maintaining ownership rights and data privacy thanks to the platform’s design. Buyers could also find the data trading platform to be a great value add as their AI may be able to be trained on more complete data as sellers may feel more comfortable allowing more parts of their data to be accessed by the AI due to the design of the data trading platform. More about the data trading platform is mentioned above and detailed in DeepBrain Chain’s whitepaper.
The DBC Token
According to DeepBrain Chain’s whitepaper, the DBC token will have a total circulation supply of 10 billion with 50% of the supply being produced through mining; 9% of the total supply were available as part of a public token sale and 6% were available in a pre-sale to professional investors and AI manufacturers; 10% were reserved for the DeepBrain Chain team along with 25% being reserved for the DeepBrain Chain foundation and ecosystem.
Within DeepBrain Chain, DBC tokens will be used as payment for activities done on the platform. For example, if someone wants to buy or train data on the platform, they will be able to conduct their transaction in DBC. Additionally, DBC will be used to fund GAS payments. With the first round of the AI Computing Package sale coming up on Aug 8, individuals and enterprises will be able to purchase GPU computing power on the platform with DBC acquired from payments in ETH.
DeepBrain Chain is creating innovation in both the worlds of AI and blockchain by creating the first blockchain-driven AI platform. The DeepBrain cloud platform already has over 200,000 users (DeepBrain Chain, 2018), and implementing blockchain into the technology could result in attracting more new users. The value proposition of creating a more secure data marketplace could be revolutionary. AI, cloud computing, and blockchain are all sectors that have experienced immense growth over recent years, and are projected to continue to grow in the future — DeepBrain Chain is applicable to all 3 of these sectors.
In addition to creating innovative technological advances, DeepBrain Chain consists of a team of experienced professionals with backgrounds across multiple areas, including AI and major tech companies. According to DeepBrain Chain’s whitepaper, they have a partnership with NEO, which was a very fast-growing cryptocurrency in 2017. As of the time of writing, NEO has the 13th-highest cryptocurrency market cap (CoinMarketCap, 2018).
Considering that DeepBrain is already a successful cloud AI platform, the team behind DeepBrain Chain yields extensive experience, and alongside with a partnership with the renowned and influential blockchain project, NEO, DeepBrain Chain has a lot of potential to change the landscape surrounding artificial intelligence.
Author: Mia S.
Revisor: Jay Jie
 DeepBrain Chain. (2018). Retrieved from https://www.deepbrainchain.org/. Accessed July 26, 2018.
 DeepBrain Chain Whitepaper. (Version 1.1.0). Retreived from: https://www.deepbrainchain.org/assets/whitepapers/DeepBrainChainWhitepaper_en.pdf. Accessed July 26, 2018.
 DBC Sale. (2018). Retrieved from https://www.deepbrainchain.org/scm.html. Accessed July 26, 2018.
 CoinMarketCap. (2018). Retrieved from https://coinmarketcap.com/. Accessed July 26, 2018.
 BlockchainConnect Conference. (n.d.). [DeepBrain Chain logo]. Retrieved July 27, 2018, from http://goblockchainconnect.com/conference201801/
DISCLAIMER: This is not sponsored. Information provided in this article does not constitute investment advice, financial advice, or any other sort of advice. You shall not treat any of the website’s content as such. Please do your own due diligence and consult financial advisory before making any investment decisions. Cypher Core will not be held responsible for the investment decisions you make based on the information provided in this article.