AI Blockchain raises $1M and successfully launches token to Ethereum
- Website: https://hypertensor.org
- Whitepaper: https://hypertensor.org/whitepaper.pdf
- Discord: https://discord.gg/pRCuCg7q
- Twitter / X: https://x.com/hyper_tensor
Previous Articles:
- Introducing Hypertensor — Decentralized Machine Learning
- Presale For An AI L1 Blockchain Is Now Live, By Hypertensor
- AI Blockchain Raises $1m — A Sold-Out Presale By Hypertensor
Token:
- Etherscan: https://etherscan.io/token/0x16A3543Fa6B32CAC3B0A755f64A729E84f89a75C
- CoinGecko: https://www.coingecko.com/en/coins/wrapped-hypertensor
Intro
The Hypertensor Blockchain sold out its presale in January 2024 selling out for a total of ~$1M in DAI (a stablecoin pegged to the US dollar). Now in May, the wrapped version of the Hypertensor token that resides on the Ethereum blockchain under the symbol $TENSOR is live on Uniswap.
What Is Hypertensor
Hypertensor is an AI layer 1 blockchain that decentralizes, incentivizes, and democratizes artificial intelligence. The blockchain interfaces with subnets that are peer-to-peer decentralized deep learning applications. These subnets decentralize the hosting of open-source machine-learning models like LLMs, image diffusers, artificial intelligence agents, and more AI categories.
Subnets & Incentives
In each subnet, each peer hosts a subset of a model alongside other peers hosting other pieces of the model, together hosting an entire model. These peers are responsible for hosting models and generating consensus amongst peers.
During each epoch, each peer must submit consensus data on each eligible peer in the hosting model. This consensus data ensures that peers interfaced with the blockchain are verified to be hosting a model, and are also submitted with a score based on computational contribution towards the model. Once scores are submitted to the blockchain, all peers must agree on the scores given to each peer based on their computational contributions as a server hosting a model. The scores given to a peer by each other peer are then put through an interquartile algorithm that is followed by the averaging of the resulting data. Peers who may be dishonest with scoring peers and exceed the allowed delta from the average score are at risk of being removed as a peer and having their stake slashed.
To read more about Hypertensor, the Whitepaper is a great place to start!
- Website: https://hypertensor.org
- Whitepaper: https://hypertensor.org/whitepaper.pdf
- Discord: https://discord.gg/pRCuCg7q
- Twitter / X: https://x.com/hyper_tensor
References
Hypertensor Testnet v1.0 Whitepaper: https://hypertensor.org/whitepaper.pdf
