【imgnAI】 Generative AI Platform Leveraging Blockchain / Providing Image Generation AI and AI Companions / Issuing Unique Token $IMGNAI / Raised $1.6 Million in Seed Round
What will be the future of decentralized generative AI tools?
Good morning.
I am mitsui, a web3 researcher.
Today, I researched about "imgnAI".
Table of Contents
1. Overview | What is imgnAI?
-What is "Nai"?
-What is "Naifu"?
2. Tokenomics | AI Tools based on $IMGNAI Token
-Details of "$IMGNAI"
-Differences from conventional AI tools
3. Evolution and Prospects | Starting from December 2022, rapid expansion, raising $1.6 million in seed round
4. Analysis | What does the future hold for decentralized generative AI tools?
Overview | What is imgnAI?
"imgnAI" is a generative AI platform that utilizes blockchain. Currently, it provides an image generation AI called "Nai", and preparations are underway to also provide an AI companion called "Naifu".
◼️ What is "Nai"?
Nai is an image generation AI bot that can be integrated as a bot on Discord and Telegram. It allows you to generate images with a similar user experience to Stable Diffusion and midjourney. There is also a web application available, which allows image generation through the web.
The usage method on platforms like Discord is the same as existing AI tools. To use it, you need to introduce a bot and input prompts to generate images on Discord. There is also a command available to modify and regenerate the generated images based on the four options provided.
Some rather unusual commands (I wonder if existing AI tools also have them?) include Remix, Chaos, Private, and Lock.
Remix: Adds randomly generated prompts to mix art styles.
Chaos: Adds a large number of randomly generated prompts to fundamentally change the art style.
Private: Prevents other users from viewing the input image as a thumbnail.
Lock: Prevents other users from generating that variation.
In addition, Nai offers a wide range of custom-trained models created by the imgnAI team using Stability AI's open-source Stable Diffusion.
gen: The most advanced model with astonishing image quality and nearly infinite art styles. It can create everything from photorealism to complex animations and digital renderings.
Supra: Offers a wide range of digital art styles with a touch of realism.
ani: Trained on classic anime and generates anime images.
nai: Trained based on the most popular anime art styles. Input detailed or basic prompts to generate beautiful anime-style art in an almost infinite range.
Hyper: Fills the gap between digital rendering and photorealism, ranging from detailed 3D renderings to lifelike realism.
art: Designed to generate highly detailed images with advanced artistic sensibilities.
Users can choose their preferred model from these options to generate images.
◼️ What is "Naifu"?
Naifu is an AI companion. It responds to user questions in a chat format, similar to ChatGPT. However, the responses are not impersonal but rather personalized to each user, with each AI having its own personality. Details are not yet known as it is still in the early access recruitment phase, but it is likely that users can choose the appearance and personality to some extent, similar to an AI friend, and it will continue to personalize over time.
Tokenomics of Tokeconomics | AI Tools Based on $IMGNAI Tokens
At first glance, it may seem like just another typical generative AI tool, but "imgnAI" issues its own unique token, "$IMGNAI".
◼️ Details of "$IMGNAI"
A total of 1 billion tokens will be distributed according to the following tokenomics:
When I checked CoinMarketCap, the current price was 2.63 yen.
As a utility for "$IMGNAI," the service "imagnAI" can be used to access the premium tools of imagnAI, as well as to gain access to the early beta version by holding a certain amount. Eventually, it can also be used as a governance token when the operation is decentralized.
The premium features include high resolution, custom image ratios, access to premium-only image generation models, watermark removal, and access to NSFW functionality.
Additionally, there is a feature in development where the generated images can be minted directly as NFTs. It is mentioned that a small amount of "$IMGNAI" will be required for this NFT generation.
The tokens used for accessing the premium features and minting NFTs will be burned, leading to a gradual decrease in the supply.
◼️Differences from General AI Tools
imagnAI mainly highlights two differences from general AI tools such as Stable Diffusion.
Firstly, it mentions the presence of an in-house AI development team. Many AI tools only offer API integration, which limits the ability to develop proprietary models or make detailed modifications. With imagnAI's in-house AI development team, it is possible to generate models trained specifically on the artwork of a particular NFT project, for example.
The second difference lies in the utilization of blockchain and a decentralized system. By primarily using tokens for payments, imagnAI can smoothly handle global transactions. Traditional Web 2.0 AI businesses often have a single point of failure due to restrictions imposed by national power balances or payment providers. This can result in limitations on the images that can be generated and the service becoming unavailable due to platform intentions or bugs. imagnAI aims to solve this problem and create a world where people can freely use AI tools for generation without censorship, through decentralized operation and payment.
Transition and Outlook | Started in December 2022, Rapid Expansion, Raised $1.6 Million in Seed Round
imagnAI was released on December 9, 2022. At that time, there were very few consumer-oriented AI projects in the web3 industry. It had minimal recognition with only 11 followers at the time of release. However, it quickly gained attention after the release and within three months, the number of followers increased to 11.4K
And on January 26, 2024, we announced a seed round of funding. This round was led by Hack VC, with participation from dao5, Rana Capital, Selini Capital, West Ham Capital, Motus Capital, and strategic angels.
This funding will be used to enhance imgnAI's infrastructure capabilities, accelerate product development, and expand marketing activities. Additionally, this funding will be used to support the launch of a product called "Naifu," a "digital companion." This product is designed to provide high creative freedom and high-performance text and image responses compared to alternative products in Web2.
Analysis | The Future of Decentralized Generative AI Tools
Finally, let's discuss.
AI×web3 tools have been researched for some time now, and "imgnAI" is one of the oldest services that focuses on generative AI for image generation and AI companions.
As I mentioned in previous research, I believe that AI and web3 are highly compatible. Generated images can be converted into NFTs, and the personality of AI companions can be turned into NFTs for buying and selling. NFTs are tools for buying and selling digital data, so it is obvious that they are well-suited for "generative" AI, where anyone can generate digital data.
However, the question is whether web3×AI tools can surpass existing tools such as ChatGPT and Midjouney. While "imgnAI" is undoubtedly useful, to succeed in the business competition, it is necessary to replace existing tools, so it is important to understand where its strengths lie.
I haven't deeply researched AI, so there is also a strong element of intuition here, but when I look at AI tools, platforms, and the evolution of AI, I feel that ChatGPT and Stable Diffusion exist as foundational layers, and application layers utilizing their APIs are emerging.
ChatGPT has also released GPT Stores, allowing monetization of applications that use ChatGPT. This enables the emergence of applications specializing in specific areas and improves the versatility of AI. Stable Diffusion is similar, and "imgnAI" is also built using models from Stable Diffusion. With its open-source nature, application development utilizing these data is progressing.
To give an analogy, it is similar to the relationship between chains and Dapps. Therefore, "imgnAI" should not compete against the infrastructure layers of ChatGPT and Stable Diffusion but rather compete against chatbots and Character .ai that utilize them, as well as specialized image generation AI tools.
Considering this, my personal conclusion is that the expansion of web3×AI generation tools is linked to the web3 market. Rather than people who want to use AI tools gravitating towards "imgnAI," it is more about "imgnAI" being an option for those in the web3 market who want to utilize AI.
The benefits of decentralization, such as the absence of a single point of failure and decentralized operation, are honestly not easily felt by general users who use it. So, as web3 adoption and the concept of NFTs become more widespread, and when it becomes possible to buy and sell images as NFTs, sell AI companion personalities as NFTs, and achieve interoperability between various AI tools, I have a feeling that competition between existing AI tools and web3×AI tools will begin. Currently, competition has not even begun, and it feels like we are still in the stage of improving products and testing hypotheses for governance in line with the expansion of the web3 market.
Personally, I believe that there is a great potential in being able to sell NFTs from AI-generated images and personalities, so I expect significant growth in generative AI and web3 services. However, it seems that a little more time is needed... that's the current analysis.
That concludes the research on "imgnAI"!
Official links: Website / Twitter
Disclaimer：I carefully examine and write the information that I research, but since it is personally operated and there are many parts with English sources, there may be some paraphrasing or incorrect information. Please understand. Also, there may be introductions of Dapps, NFTs, and tokens in the articles, but there is absolutely no solicitation purpose. Please purchase and use them at your own risk.
About us
A web3 newsletter delivered in five languages around the world. We deliver various articles about web3, including project explanations, news and trend analysis, and industry reports, every day. You can subscribe for $8 per month ($80 per year) and receive research articles that took 100 hours to complete every day.
Author
mitsui
A web3 researcher. Operating the newsletter "web3 Research" delivered in five languages around the world.
Contact
The author is a web3 researcher based in Japan. If you have a project that is interested in expanding to Japan, please contact the following:
Telegram：@mitsui0x
*Please note that this newsletter translates articles that are originally in Japanese. There may be translation mistakes such as mistranslations or paraphrasing, so please understand in advance.