Introduction

This year, the AI industry has attracted significant attention. The projects within this field can be categorized into two types. The first type includes projects that were originally unrelated to AI but have leveraged the “AI concept,” such as Worldcoin, co-founded by OpenAI CEO Sam Altman. The second type involves projects that are closely related to AI technology.

This article will be divided into two parts: the first will cover “AI concept projects” that have capitalized on the AI craze, and the second will inventory the top ten high-market-value AI projects.

AI Concept Projects

What is Worldcoin?

Worldcoin is a project that aims to implement UBI (Universal Basic Income). It offers World ID, an identity verification network that operates under the premise of protecting privacy. Unlike KYC, its fundamental principle is to allow anyone to prove they are human without providing personal information such as phone numbers, ID numbers, or names. They ultimately chose to develop a hardware device called Orb for iris verification.

The design and development of Worldcoin are supported by partners like Tools for Humanity (TFH) and are overseen by the Worldcoin Foundation. Moreover, the patent for the Orb hardware design used for iris verification belongs to the foundation.

The project’s most notable co-founder is Sam Altman, the co-founder and current CEO of OpenAI. Top-tier institutions led by a16z have invested in Worldcoin, with financing exceeding $240M, making it a star project.

The adoption rate of Worldcoin is continuously growing, with over 4.35 million users, all of whom are biometrically verified and unique humans. Worldcoin is currently integrating with major Web2 platforms such as Minecraft, Reddit, Telegram, and Shopify.

Past articles published by Gate have provided detailed introductions to Worldcoin, its working principles, and other projects also developing ZKID technology.

Worldcoin and the Concept of AI

$WLD benefits from the halo of its co-founder Sam Altman, who believes that the advent of AGI (Artificial General Intelligence) will replace a large number of human jobs. Therefore, before that happens, there needs to be a project that realizes the ideal of unconditional basic income, redistributing the wealth brought about by AI to the world. The relationship between Worldcoin and AI is built on the possibility of OpenAI and future AGI.

$WLD has a very high total supply (10B), making Worldcoin’s FDV reach 94B, ranking fourth after BTC/ETH/SOL. With the selling pressure brought by token unlocking (see the figure below), it was once thought to affect the token price; however, the current circulating supply of $WLD is only 1.53%. Following the release of the example video of Sora by OpenAI on February 15th, which showcased leapfrog progress in the field of AI-generated text to image, the price of $WLD doubled within a week, showing strong market consensus on Worldcoin.

Source: Official document, 15-year $WLD unlocking supply schedule

What is NEAR Protocol?

NEAR is a Layer1 public chain that existed in the last bull market, aiming to be one of the “Ethereum killers.” Its key differentiation as a public chain is sharding technology, which is considered one of the key paths to improving the performance of public chains, with an average TPS of about 300; Ethereum initially adopted the sharding technology route but later turned to Layer2 solutions due to its technical complexity.

The current roadmap progress of NEAR is as follows: The NEAR team decided to initiate Nightshade Sharding on January 30, 2024, which is the second phase of the sharding test network, simplifying the tasks for validators. The second phase of sharding technology is likened to the community upgrading to a super-high-speed transportation system and communication tools. By introducing “stateless verification” technology, whereas each region (shard) had to maintain detailed ledgers of every transaction and data, they can now obtain this information from a centralized “library,” thus making transaction verification more convenient and faster.

NEAR has several notable R&D directions: zkWASM in collaboration with Polygon, the fast settlement sorter in cooperation with EigenLayer, and the independently constructed NEAR DA, among others. As for chain abstraction, it is still in the development stage. The current NEAR supports sharding technology and EVM compatibility, focusing on carbon-neutral Gas Fees and a user-friendly DApp experience as its core competitiveness.

NEAR’s financing situation shows its favor in the capital market, with the most recent financing in April 2022, amounting to $350 million and covering many well-known investment institutions. For an introduction to the NEAR project, refer to the Gate article NEAR Research Analysis Report.

NEAR Protocol and AI Concepts

At the highly anticipated 2024 NVIDIA GTC Conference (March 17–March 21), NEAR Protocol’s co-founder/CEO Illia Polosukhin is set to speak on the panel “Transforming AI,“ alongside Nvidia CEO Jensen Huang. This announcement saw the price of NEAR’s token surge by more than 100% within a week.

Illia Polosukhin has a significant background in AI, having led a team at Google Research that developed deep learning and NLP technologies. He is also an open-source contributor to TensorFlow.

The relationship between NEAR Protocol and AI was hinted at in an article by Vitalik Buterin on January 30, 2024. Buterin discussed a dilemma: some users are concerned about bias or deceit within the system and thus demand AI to be open-sourced. However, for greater security, there’s a need to maintain AI as a “black box.” In this dilemma, utilizing blockchain technologies like zero-knowledge proofs or multi-party computation (MPC) to create what he calls “trustworthy black box AI” is very meaningful. Vitalik mentioned that NEAR Protocol is developing some technologies and has made this vision a core goal.

However, NEAR Protocol has not yet announced any actual innovative results related to AI technology. The price increase is mostly due to the market’s expectations for the AI products NEAR Protocol might launch. Therefore, it is still considered an AI concept project, worth continuous attention.

What is Arkham?

Arkham integrates over 3.5 million wallet tags and 200,000 entities, connecting wallet addresses to their actual owners (entities) through algorithms and weights set by its proprietary ULTRA system.

Arkham was founded in 2020 by Miguel Morel, who is also a co-founder of the Reserve Protocol.

Funding situation of Arkham: On July 18, 2023, it raised $2.5 million in a Public Sale; on November 15, 2023, Binance Labs invested an undisclosed amount.

Tokenomics of Arkham: The $ARKM token has three main functions: it serves as the currency for Arkham Intel Exchange and the Intel-to-earn economy; it creates an incentive system through rewards/discounts to accelerate the adoption of Arkham; and it acts as the governance token for the Arkham Intel Exchange. The initial total supply of ARKM is 1 billion tokens, distributed as follows:

Source: Official website

Arkham and AI Concept

One of the most noteworthy aspects of Arkham is the personal investment and favor it has received from Sam Altman and Peter Thiel. It has developed many AI-based internal tools and algorithms for entity analysis and “de-anonymizing” on-chain data, such as identifying the real users behind multiple wallets.

Arkham and Worldcoin share similar controversies. Some believe that de-anonymization distorts the original intention of blockchain’s anonymity and emphasis on privacy. The opposition is not necessarily against the biometric data used by Worldcoin, but as blockchain adoption increases, if many projects integrate World ID as a verification method for anti-sybil purposes, it may completely change the working mode of many exploit studios, script and bot users (on-chain scientists), and institutions and market makers controlling numerous wallets, leading to a redistribution of interest structures.

In summary, Arkham is an on-chain data analysis platform integrating various machine learning algorithms. Due to the noise problem with on-chain data, AI+on-chain data analysis is still in its early stages. Similar products mainly rely on adjusting the weights of machine learning algorithms to achieve their primary functions. In practice, they still rely heavily on manual in-depth analysis, making it difficult to explore data directly with AI. However, Arkham’s ecosystem encourages on-chain analysis and exploration projects, offering many bounty programs with rewards on the Arkham Intel Exchange, motivating researchers and on-chain detectives.

Other: Data Infrastructure Projects

The three most crucial elements in the AI race are computing power, data, and models. Although many projects do not focus on aggregating or providing solutions for AI model training data, they are still significantly influenced by the AI hype. The Graph ($GRT) is an indexing protocol for querying data from Ethereum and IPFS. At the end of 2022, with the release of ChatGPT, it gained considerable momentum due to its role in data infrastructure, riding the wave of AI enthusiasm, and its market value currently ranks #35. Notable projects in decentralized data storage include Filecoin (FIL), with a market value ranking of #26, which is an incentive network based on IPFS distributed storage, and Arweave (AR), with a market value ranking of #52, offering a solution for permanent storage.

AI Projects

The Top Ten AI Projects by Market Capitalization

The table below includes the top ten AI projects directly related to the business and AI technology, ranked by market capitalization:

Source: Author

Render

The Render network is a high-performance distributed GPU rendering network, leveraging industry-leading OTOY Inc. software to aggregate GPU suppliers and those in need of GPU resources, thereby creating a rendering power marketplace.

The working principle of the Render network is as follows: owners connect their GPUs to the rendering network to receive and complete rendering jobs using OctaneRender; users send $RNDR tokens to individuals executing the rendering work, with OTOY receiving a small portion of $RNDR to facilitate transactions and operate the rendering network.

Render was founded by Jules Urbach in 2016. During its ICO in 2018, Render raised $1.16M; in December 2021, it secured $30M in a strategic funding round led by Multicoin Capital.

Since October 2017, a total of 117,843,239 $RNDR tokens have been allocated, with 536,870,912 $RNDR tokens already sold publicly. The total circulating supply is 156M, with a total supply of 522M and a maximum supply of 536M. 25% of the tokens were sold, 10% were used for reserves, and 65% were entrusted for custody (for supply/demand recirculation into the network). The $RNDR ICO in October 2017, along with private sales from January 2018 to May 2018, priced each $RNDR at $0.25. Additionally, the RNDR network plans to purchase up to 4.5 million RNDR tokens on the Probit exchange to flow backe to creators and new users within the RNDR ecosystem. So far, 3.2M tokens have been acquired from ProBit, with an average price of approximately $0.042.

Currently, Render, with the highest market value among AI projects, is considered the Nvidia of cryptocurrencies. It has recently been recognized multiple times by the Web2 world, such as being integrated into Nvidia Omniverse and being mentioned in last year’s Apple WWDC 2023 keynote speech.

Bittensor

Bittensor is an open-source protocol, and its TAO network can be utilized to construct decentralized, blockchain-based machine learning networks. Machine learning models can collaborate within the network to perform training tasks, earning token rewards in the TAO network based on the value of the information. The TAO network also provides external access, allowing users to extract information from the network while offering high customizability. Initially planned to become a parachain on Polkadot, concerns about the development pace of Polkadot led to the decision to utilize its own independent Layer 1 blockchain. In practice, the TAO network itself does not execute on-chain machine learning tasks; instead, it acts as an on-chain oracle or becomes a verification network connecting off-chain machine learning nodes. Bittensor’s ultimate goal is to create a pure market for artificial intelligence, an incentivized environment where buyers and sellers of valuable information (model training tasks/data) can interact in a trustless, open, and transparent manner. This market offers a permissionless platform, in stark contrast to closed models like OpenAI or Google Gemini.

Bittensor was founded by Jacob Robert Steeves and Ala Shaabana in 2021. The team has not disclosed any financing amounts, but public data shows Bittensor underwent an accelerator incubation on January 1, 2021, and an early-stage Series A financing on May 12, 2021. In terms of fairness, Bittensor ($TAO) did not conduct a token pre-mining, and no proportion was allocated to investors, market makers, the team, or advisors, thus avoiding the selling pressure from token unlocking, making it a fairer project for retail investors. Similar to Bitcoin, the total maximum supply of $TAO tokens is capped at 21,000,000, with a halving occurring every 10,500,000 blocks, totaling 64 halving events. Currently, 7200 new $TAO tokens are issued daily, evenly distributed between miners and validators, hence $TAO is referred to as a “mining coin.”

Bittensor has been listed as number one in the top ten market cap list for a period; it was also mentioned by Vitalik as one of the cases that “choose to use cryptographic incentives to motivate the creation of better artificial intelligence” instead of adopting fully encrypted complex schemes without the use of cryptography.

Theta Network

Theta Network is a blockchain network focused on media, artificial intelligence (AI), and entertainment. Its goal is to increase revenue for existing video or streaming media platforms through infrastructure improvements and to reduce the costs of content delivery network (CDN).

Theta Network is not a decentralized version of YouTube or Twitch but rather a foundational blockchain and decentralized storage/delivery infrastructure. In practice, Theta Network consists of two complementary subsystems: the Theta blockchain and the Theta edge network. The Theta blockchain offers payment, rewards, and smart contract functionalities, while the edge network is responsible for the storage and delivery of media assets. Theta Network features a unique multi-level BFT consensus design, consisting of enterprise validator nodes and guardian nodes operated by the community. It rewards users for sharing storage and/or bandwidth through a P2P market and supports Turing-complete smart contracts, enabling the creation of various Web3 dapps on Theta Network. Smart contracts on Theta Network allow for fully digitalized ownership of items, innovative payment consumption models, and transparent royalty distribution. The Theta blockchain can process 1000 transactions per second with a nearly zero carbon footprint.

Theta Network employs a dual token design. The Theta token (THETA) is the governance token of the Theta protocol. It is used for staking as a Validator or Guardian node, contributing to the block production and protocol governance of the Theta network. By staking and operating nodes, users earn a certain amount of new TFUEL. The supply of THETA is fixed at 1 billion and will never increase. Theta Fuel (TFUEL) is the operational token of the Theta protocol, used for on-chain operations such as paying edge node relays to share video streams or deploying and interacting with smart contracts.

Mitch Liu and Jieyi Long founded Theta Network in 2017. The team secured $20M in funding led by DHVC in 2017; and in May 2019, they received undisclosed funding led by Samsung Next and Blockchain.com. Theta Network also has an enterprise validator and governance council with members from Google, Samsung, Sony, Binance, and content collaboration partners like Lionsgate and MGM Studios. They have also partnered with companies like Samsung and Sony, earning recognition and adoption in the Web2 world.

In a ranking, Theta Network’s market capitalization is only surpassed by Render and Bittensor. However, it seems to have been overlooked in Coingecko’s AI project category, with the website’s project description still referring to it as a “decentralized video streaming network” without mentioning its relation to AI. Currently, Theta Network’s core edge node functionality can transform users’ personal computers into edge computing nodes, performing various tasks including AI/deep learning model training and inference, as well as video transcoding and relay. By running edge nodes, users can contribute their unused computing and bandwidth resources to earn token rewards. Theta Network’s EdgeCloud is attempting a large-scale decentralized computing project for mobile devices. From these aspects, Theta Network’s close relationship with AI is evident.

Nosana

So, what is the fastest-growing AI project this year? According to a research report by CoinGecko, Nosana ($NOS) has risen from $0.56 to $6.01 since the beginning of the year, with a price increase of over 900%, making it the fastest-growing AI project this year. Nosana is not in the top ten market cap table, falling at eleventh place with a market cap of $493 million.

Source: CoinGecko Research

Nosana provides a decentralized GPU grid that can run AI inference tasks for common models, such as LLama 2 and Stable Diffusion. It allows anyone to rent computing power without high costs or token lock-up.

The team’s roadmap outlines the key stages of development, mainly divided into five release cycles:

• Genesis: The platform’s birth, a rigorous testing and refinement phase
• Galactica: Expansion phase, opening the network to the public
• Triangulum: Focused on improving interoperability, integrating with major machine learning protocols
• Whirlpool: Supports various types of GPUs and devices
• Sombrero: Dedicated to catering to businesses of various scales by introducing financial collaboration features

Nosana was founded by Sjoerd Dijkstra in Amsterdam in 2021, with no disclosed funding. The team launched a private sale of $NOS on February 12, 2022. Currently, $33 million worth of $NOS is in circulation, with a circulating supply of 83.4M and a total supply of 100M. $NOS can mainly be purchased on Gate.io/Radium/MEXC.

Last year, Nosana migrated the project to the Solana ecosystem and shifted towards the aggregation of computing power for AI inference tasks, currently entering the second phase of the “Incentivized Public Test Grid” program.

Conclusion

This article analyzed well-known AI concept projects that have risen with the AI craze, as well as the top ten high market cap AI projects that have been deeply involved in the AI business for many years, and additionally introduced Nosana, the fastest-growing AI project this year.