On January 26th, local time, at the invitation of the organizers, Feng He, CEO of DeepBrain Chain attended the Blockchain Connect Conference and delivered a keynote speech. As one of the largest gatherings of top scientists, investors, entrepreneurs in the blockchain industry, the conference attracted attention from over all the world.
Feng He introduced that DeepBrain Chain is the world’s first AI computing platform powered by blockchain. A crossover creation that tries to solve some of the most challenging issues in the AI industry, namely: high cost and data privacy. He also gave a detailed introduction of DeepBrain Chain’s ecosystem, with AI companies at the center, interacting with the other 3 elements, i.e. miners, exchanges and data, model, algorithm and docker providers. Its token, DBC, is the link amongst all 4 elements.
Built on top of the NEO system, DeepBrain Chain creatively proposed a mining mechanism. Instead of asking miners to do meaningless computation only to lead to wasting power, reward miners for providing computing power.
Below is the full text of his speech:
"Good afternoon everyone, my name is Yong He, I’m from China. I am the CEO of DeepBrain Chain.
DeepBrain Chain Foundation is a Non-profit organization registered in Singapore, if you ask me what DeepBrain Chain is. To summarize it in one sentence, it is an AI computing platform powered by blockchain.
Let’s take a moment to picture the AI market. PWC once predicted that by 2030, AI industry will contribute over 15.7 trillion U.S dollar to the global economy. This is more than the combined output of China and India. This shows how big the AI market is. Accenture research shows that AI has the potential to boost corporate profitability by an average of 38% by 2035. Currently the total investment in AI has already exceeded 100 billion USD. However, the cost of AI is equally as great. In fact, 10% - 30% of all AI application cost is the computational hardware cost. For a single AlphaGo system it costs 25 million USD and, uses 1,920 CPU’s and 280 GPU’s. Just one game of AlphaGo costs 3,000 USD. This is where DeepBrain Chain comes in. We empower the whole AI industry, we can bring AI companies up to 70% saving on computing cost.
DBC has 3 core features: smart nodes, high security and scalability. Different mining nodes will compete to deploy AI containers, millions of smart nodes are ready to process efficiently. The current AI processing is vulnerable to security threats and data leakage. DBC provides a secure, insulated environment where data and AI models can be stored. All DBC data is encrypted, every node is cloaked. We can give value to your data. We confirm ownership of data while allowing people to buy and sell data usage rights, thus creating a huge data market. This, combined with the right incentive mechanism, gives data providers enough reason to share data, creating a network effect. As more people share data, data will no longer be the bottleneck in the AI industry. AI processors provide powerful processing capabilities. DBC is collaborating with chip manufacturers to optimize AI processors. DeepBrain Chain’s computing neural network nodes are dynamic and will adjust according to the computational demand.
We are currently building an AI and Blockchain lab in Silicon Valley, focusing on cutting-edge technology research. The lab will help DBC achieve breakthrough in 3 technical aspects:
- Architecture wise, neural network large-scale parallel training.
- DeepLearning Algorithm, which is Communication-Efficient Learning of Deep Networks from Decentralized Data.
- The saving aspect, use Reinforcement Learning to reduce energy consumption across the DBC network.
In DBC ecosystem, AI companies are a crucial part to the system. The more AI companies get involved the more valuable DBC is. AI companies buy DBC from the exchanges, then pay DBC for computational power; they can also buy data with DBC, or models, algorithms and mirrors. Investors can invest in DBC on exchanges, this completes a closed-loop business model.
Finally, DBC wants all parties to benefit from the platform. For miners, they are rewarded for mining via tokens; for businesses, they can save up to 70% of AI processing; for data owners, data trainer and data provider are separated using smart contracts. Hence the value of the data is protected. data owners can sell their data with safe data usage rights to monetize the value of the data created, providing a new income for data owners.
From when the project was activated back in March 2017, we have built a global community in 7 months time. By Q2 this year we plan to conduct internal testing of DeepBrain Chain; and community growth to 100 thousand users in 100 countries. By Q3 this year we plan to do Alpha testing and mining. By Q4 this year we plan for the community growth to reach 500 thousand users in 100 countries. As well as Beta testing, we will also invest in 10 quality AI companies and over 100 AI companies will be running on DBC. We believe that AI will change the world, and DeepBrain Chain will help the AI industry grow stronger.
If you want to know us more, you can follow us on Twitter, or join our telegram chat room, or even send me an e-mail!"
As one of the few projects that have practical applications in the industry, DeepBrain Chain has attracted a lot of attention since its creation. Nature listed it alongside Baidu, Tencent and IflyTek as “AI Projects to Watch”. With the accumulation of more than 7 years in AI, the DeepBrain Chain team has a deep understanding of the industry. This, combined with blockchain’s decentralized and temper-proof nature, provides a perfect solution to the two most pressing issues hindering the development of AI companies.
Company name: DeepBrain Chain
Company site: https://www.deepbrainchain.org
Joya Lin: email@example.com