Trading, markets, and decentralized finance are of the most valuable use cases of AI. Now Phoenix is enabling AI-powered DeFi through AlphaNet, Hypermatrix, and SkyNet.
Will this be the start of the AI-driven decentralized finance revolution?
Love powerful customization in generative AI? Good news is that the first major upgrade will be coming to Phoenix GenAI’s Model Fine-Tuning features this week. This upgrade includes the rollout of the GenAI Depth-to-Image model, which enables users to swap out and enhance existing elements (objects, background) in an image without altering the existing style.
This upgrade will also include additional preloaded fine-tuning LoRA models for artistic style enhancement, selectable by a simple click by users. $PHB
JDI and Tensor to invest $20M to create Phoenix AI Ecosystem Fund, a strategic development fund for both accelerating growth of the Phoenix AI application ecosystem as well as investing in AI compute infrastructure hardware.
After months of meticulous R&D, today we are glad to unveil our next breakthrough innovation in AI-driven markets and trading, AlphaNet Hypermatrix. Hypermatrix is an institutional, enterprise-grade AI modeling infrastructure built on SkyNet, used for customization, advanced modeling, and optimization of AI-based systematic trading strategies.
With Hypermatrix, ecosystem partners are able to integrate and customize AI trading strategies and insights to create transformative value for their users as an AI-enabled exchange, DEX, or DeFi platform. As a partner platform, leveraging Hypermatrix’s infrastructure will enable custom-tailored tweaking based on signal/strategy type, Sharpe ratio, trading strategy, average win/loss, and across over 30+ parameters, with over 20 million AI strategy and model combinations for their users. All trading pairs currently available on Binance Futures will be supported via Hypermatrix. In essence, Hypermatrix and SkyNet will be a turnkey solution for AI compute, infrastructure, and modeling for our partners.
We would also like to use this sneak peek as an opportunity to welcome @BellaProtocol and @SYP_Protocol to the Phoenix AI Partner Ecosystem. More details regarding Hypermatrix, and our new ecosystem partner integrations, will follow shortly.
Integrations, pilots, and testing are commencing as you are reading this message. With this announcement, we are confident to say that Phoenix and AlphaNet are pushing the frontier in AI-enabled trading and DeFi.
As we have exceeded our initial node quota of 3000 units, PhoenixNode pre-registration is now closed.
Those who submitted past the 3000-unit mark will be put on the waitlist for pre-orders. Any node supply surplus due to registrants reducing or omitting the specified node count will be made available to those waitlisted and new participants in pre-order stage, which our hardware partner http://Bobber.com and JDI will provide more details in advance. We thank you for your understanding.
Phoenix GenAI Model Fine-Tuning is now online. #GenAI users are now able to adjust styles, prompt influence weight, and specific elements. A comprehensive user guide will be available early next week.
To use, simply access the PhoenixLLM bot on Telegram (SkyNet account must be binded), select “Settings” on the menu, and select “SDXL Settings”. After adjusting, go back and call GenAI to use SDXL with fine-tuning settings. Fine-Tuning will also be made available for Base Model and Text2Motion.
For beginners, we recommend playing with the LoRA’s first!
Optimization and adjustment of styles and elements in AIGC models are versatile tools for personalization and customization of the output. In a few days we will rollout 𝗚𝗲𝗻𝗔𝗜 𝗠𝗼𝗱𝗲𝗹 𝗙𝗶𝗻𝗲-𝗧𝘂𝗻𝗶𝗻𝗴, which through a series of tools and add-on models, enable users to easily make adjustments to styles and elements to enhance the model output.
Model fine-tuning tools will be made easily accessible to users via Telegram, usable with easy commands and options.
The following AI research report from Messari highlights the fact that decentralized AI compute platforms that focus on simply the number of GPU nodes in the networks is an unsustainable approach (supply-first). Growing and building AI applications that are use case-driven catered for the AI user (demand-first) is the ideal way to develop a decentralized AI compute network.
Original Source (Excerpt Included Below): https://messari.io/report/are-gpu-compute-networks-supply-or-demand-constrained
This precisely aligns with the structure and design of Phoenix’s AI Application Ecosystem and SkyNet AI elastic compute infrastructure, which focuses on value-delivery through various vertical AI applications to grow the network. From Phoenix’s native apps such as AlphaNet (trading), PhoenixLLM (knowledge), or GenAI (content), to ecosystem app partners such as Horizon Protocol (AI DEX) and W3Goo (AI Search), delivering readily accessible use-case value for end users is the core tenet of our platform.
𝗧𝗵𝗲 𝗣𝗵𝗼𝗲𝗻𝗶𝘅 𝗧𝗿𝗶𝗳𝗲𝗰𝘁𝗮, our approach for AI value-creation, consists of 3 pillars: AI Alpha (market insights, trading), AI Insights (predictive analytics, enterprise apps, AI search), and AI Productivity (LLM, AIGC, automation).
Phoenix SkyNet AI ecosystem adoption steadily rising - today GenAI processed over 700 images and Text2Motion requests. Total GenAI has just surpassed 10,000 calls across Base, SDXL, and Text2Motion models.
PhoenixNode pre-registrations are scheduled to begin at the end of February. If you aren't familiar with PhoenixNode, review our @decryptmedia article & announcement here https://decrypt.co/208979/phoenix-partners-with-largest-helium-hardware-vendor-to-launch-ai-compute-miner
In short, pre-registration is used to gauge and approximate hardware supply preparations for our hardware and #DePIN partner http://Bobber.com / JDI for the first wave of pre-orders. It is also used to collect other preliminary metrics such as average node count per operator and colocation preferences.
For more information on pre-registrations and other things to expect for the PhoenixNode, please refer to the following FAQ: https://phoenixnode-faq.phoenix.global
This is an important technical milestone for the #generativeAI industry. Phoenix has already been experimenting with #texttovideo capabilities on GenAI starting with our Text2Motion GIF generation model released earlier this month. https://www.theverge.com/2024/2/15/24074151/openai-sora-text-to-video-ai We expect to continue rolling out additional @Telegram-integrated text-to-video features, models, and optimizations on $PHB 's GenAI.
Phoenix SkyNet serves as a next-generation AI elastic compute platform enabling users and developers to use (inference), train, and deploy a variety of AI and machine learning models without having to worry about underlying compute and scaling infrastructure.
With over 3000 active AI jobs, and 3 native AI apps running on SkyNet (not including 3rd party apps building on SkyNet), Phoenix focuses on accessible and value-driven AI that users and developers can access via various interfaces. This is namely through our API/SDK for developers, SkyNet Control Panel (codeless training and deployment), or through our ready-to-use app ecosystem, including features integrated with Telegram (and soon Discord).
It gets better – after the release of PhoenixNode users will be able to manage and contribute multiple compute nodes, as well as mine rewards, all through SkyNet.
Another Phoenix GenAI upgrade is here! In addition to existing features, a new model Text2Motion has been added, enabling users to create AI-generated animations from text. In this release of Text2Motion, outputs are available in GIF, and will be expanded to other formats as the model continues to train and optimize.
To use, simply call GenAI within PhoenixLLM in @telegram , and select Text2Motion.
GenAI is gearing up to be one of the gateways of mass adoption for SkyNet, our AI Elastic Compute Layer.