Decentralization Intelligent Network in the Wave of AI Revolution

With the rapid development of artificial intelligence technology, we are entering a data-driven new era. Breakthroughs in fields such as deep learning and natural language processing have made AI applications ubiquitous. The emergence of ChatGPT in 2022 ignited the AI industry, followed by a surge of AI tools such as text-to-video and automated office applications. The market value of the AI industry has also risen accordingly, and it is expected to reach 185 billion dollars by 2030.

However, the AI industry is currently mainly monopolized by a few tech giants, which has led to a series of challenges such as data centralization and uneven distribution of computing power. At the same time, the decentralization concept of Web3 provides new possibilities for addressing these issues. Against this backdrop, some high-quality Web3+AI projects have emerged.

Among them, Bittensor seizes the opportunity of the Web3 + AI algorithm direction, and builds an AI algorithm platform with its own screening competition mechanism through the competition and incentive mechanism of blockchain, in order to retain the highest quality AI projects. This innovative attempt provides new ideas for reshaping the current AI development pattern.

Bittensor: Pioneer of Decentralization AI Network

Bittensor is a decentralized incentivized machine learning network and digital goods marketplace. It has the following features:

1. Decentralization: Runs on thousands of distributed computer networks, avoiding the problem of data centralization.

2. Fair incentive mechanism: Network rewards are proportional to contributions.

3. Machine Learning Resource Sharing: Providing services for all individuals in need of machine learning computing resources.

4. Diversified digital goods market: tradable data in various forms.

Unlike many overvalued VC projects, Bittensor is a fairer, more interesting, and meaningful geek project. Its development history is as follows:

• 2021: Project launched, built blockchain using Substrate framework.
• 2022: Released the Alpha version of the network, validating the feasibility of Decentralization AI. Introduced Yuma consensus.
• 2023: Release of the Beta version, introduction of the TAO token economic model.
• 2024: Utilize DHT technology to improve data storage efficiency, expand subnets and the digital goods market.

The total supply of Bittensor's token TAO is 21 million, halving every four years, distributed via a fair launch method. Currently, a block is generated every 12 seconds, with each block rewarding 1 TAO. These rewards are distributed to various subnets and their members based on contribution.

Currently, the total number of accounts on the Bittensor network exceeds 100,000, with non-zero accounts reaching 80,000. The TAO token has seen a maximum increase of several dozen times in the past year, with a current market value of 2.278 billion USD and a coin price of 321 USD.

Bittensor's Subnet Architecture

The Bittensor protocol is a Decentralization machine learning protocol that supports network participants in exchanging machine learning capabilities and predictions. The network is composed of multiple subnets and employs a survival of the fittest mechanism.

Subnets are the most important components of the Bittensor network and can be seen as a segment of independently running code that establishes unique user incentives and functionalities. Currently, there are a total of 45 subnets, excluding the root subnet. It is expected that from May to July 2024, the number of subnets will increase from 32 to 64.

The subnet includes three roles: subnet owner, miner, and staking validator.

• Subnet owner: provides the underlying code and sets up the incentive mechanism.
• Miner: Iterative optimization of code, providing computing services.
• Validator: Evaluate subnet contributions and earn staking rewards.

Subnetwork emission is the reward distribution mechanism in the Bittensor network. Generally, 18% is allocated to subnetwork owners, 41% to validators, and 41% to miners. The higher the subnetwork emission, the higher the ranking.

There is a 7-day immunity period after subnet registration. The initial registration fee is 100 TAO, and the price doubles for subsequent registrations. When all subnet positions are filled, the subnets with the lowest emissions and no immunity period will be eliminated.

Bittensor's Consensus and Proof Mechanism

The Bittensor network employs various consensus mechanisms and proof mechanisms, among which the most distinctive are the Proof of Intelligence ( PoI ) mechanism and Yuma consensus.

The PoI mechanism is Bittensor's original verification and incentive mechanism. Miners prove their contributions by completing smart computing tasks, which are assigned and scored by validators. This ensures the security of the network, the quality of the data, and the efficient use of computing resources.

Yuma consensus is the core consensus mechanism of Bittensor. Validators score based on task completion and input the Yuma algorithm. The algorithm assigns different weights based on the amount staked by validators and filters out outliers, ultimately deriving a comprehensive score and distributing rewards.

Yuma consensus follows the principle of data unawareness, protecting data privacy and security. At the same time, it adopts a performance-based reward mechanism to ensure efficient and high-quality resource utilization.

In addition, Bittensor also introduces the MOE( mixture of experts) mechanism, integrating multiple expert-level sub-models to achieve better operational results. Under the Yuma consensus, validators can score and rank the expert models, incentivizing continuous optimization of the models.

Bittensor Subnet Ecosystem

Currently, Bittensor has 45 registered subnets, of which 40 have been named. Due to the introduction of an elimination mechanism, poorly performing subnets will be eliminated in the long run, retaining high-quality projects.

Apart from the root subnet, subnets 19, 18, and 1 are the most关注, with emission shares of 8.72%, 6.47%, and 4.16%, respectively.

Subnet 19 Vision focuses on Decentralization image generation and inference, providing access to open-source LLM and image generation models. Currently, the total 24-hour node revenue is approximately 627.84 TAO.

Subnet 18 Cortex.t is dedicated to building a cutting-edge AI platform, providing text and image responses through APIs. Currently, the total earnings of the nodes in the last 24 hours are approximately 457.2 TAO.

Subnet 1 is the earliest text generation subnet of Bittensor, which has faced skepticism but still maintains a high emission volume.

From the perspective of model categories, in addition to generative models, there are also various types such as data processing and trading AI. In terms of return and risk, successful operating nodes yield considerable returns, but new nodes face significant competitive pressure.

Future Outlook

1. The Web3+AI sector is expected to maintain market enthusiasm for a considerable period in the future.

2. The Bittensor project has a unique architecture, combining technical and market support, with broad development prospects.

3. The subnet architecture lowers the barrier for AI teams to enter the Decentralization network, which helps attract more quality projects.

4. With the increase in the number of subnets, it is necessary to be wary of the influx of obscure projects and the risk of yield dilution.

Overall, Bittensor, as a pioneer of Decentralization AI networks, provides an innovative model for the integration of Web3 and AI, which is worth keeping an eye on for its development trends.