What is Bittensor (TAO)?
Bittensor is an open-source, peer-to-peer, decentralized, and permissionless machine learning network.
Bittensor (TAO)’s whitepaper starts with the premise that artificial intelligence has become a commodity: it is ‘extensively mined from data’, ‘monetarily valuable’, ‘transferable’, and ‘generally useful’. By this logic, Bittensor’s whitepaper states that artificial intelligence, a new type of commodity, needs a new type of market to help produce it efficiently.
Bittensor is the proposed solution - a network described as a collectively-run, decentralized market where ‘intelligence is priced by other intelligence systems peer-to-peer across the internet’.
What does that mean?
Bittensor invites users to submit their own machine learning models to the network. These are then tested and trained by other neural networks on Bittensor, and ranked by value. Ranks are recorded on Bittensor’s digital ledger. You can think of it almost as a judgment-free way of ranking artificial intelligence systems: AI models on Bittensor are ranked for their efficiency and accuracy, regardless of the task they are being trained for.
High-ranking peers are rewarded with additional weight in the network.
Bittensor’s whitepaper outlines the benefits of this system compared to centralized alternatives as being:
- The ability to reward AI models which can applied to a larger set of objectives
- More esoteric or legacy systems can be monetized for their unique value
- Researchers can monetize their intelligence work, and consumers can directly purchase it
What makes Bittensor (TAO) different?
According to co-founder Ala Shaabana’s Medium post, ‘The Future of AI is Decentralized’, (first published in February 2021), the issues with centralized AI research and production are:
- The research is non-collaborative. AI models are siloed, and cannot share knowledge with each other. ‘If any one model learns something that may be valuable to other models, it cannot share it with them to speed up their training’.
- It is wasteful. The intelligence produced by powerful models such as GPT ‘is always lost’, meaning ‘users have to re-train these models on their own systems to replicate them’.
- The barrier to entry is too high. Researchers with low computational resources cannot contribute to the development of large neural networks - conducting the ‘necessary experiments would be too expensive’.
Bittensor claims to solve these issues by being a permissionless, decentralized, and incentivized network (like Bitcoin). Bittensor has positioned itself as :
- Collaborative. By being open-source, all knowledge within the network is shared.
- This also makes Bittensor a greener alternative to its centralized competitors. Knowledge-sharing removes the need for AI models to be completely re-trained.
- Anyone can participate in the network, and will be rewarded for their contributions.
What does the TAO token do?
TAO is a digital token which functions as currency within the Bittensor network. TAO is required to purchase access to machine learning models on the network, to reward validators and miners, and to reward researchers who contribute new machine learning models to the network.
TAO follows the same tokenomics as Bitcoin. TAO has a maximum supply of 21 million tokens, and like Bitcoin, has a four-year halving cycle. For a visual representation of the token’s emission over time, check out Taostats.
Brief project background
Bittensor was founded in 2019. The network and TAO token launched in 2021. TAO was distributed via fair-launch, meaning no tokens were pre-mined.
Where Bittensor (TAO) sees itself going
Bittensor hopes its network will eventually be used by other AI companies to validate the quality of their machine learning algorithms.
They also foresee companies of all sizes leveraging, and paying for, AI models which will only be available on Bittensor.
Features you can try out now
To better understand how Bittensor’s technology can be leveraged, you can try out some of the applications already built on the network.
- Chat with Hal - A personal AI assistant
- Reply Tensor - Produces AI-generated Twitter responses
- Tao Studio - powered by Bittensor’s testnet for text-to-image generation
Want to dive deeper into Bittensor?
For more information about the project, and to get a flavor of how the network works and how it’s doing, check out the following resources:
About the founders
Bittensor was founded by Ala Shaabana and Jacob Robert Steeves. The project is supported by the Opentensor Foundation and individual contributors.
Ala Shaabana holds a PhD in Computer Science from McMaster University. He was an Assistant Professor at the University of Toronto from 2020 to 2021.
Jacob Robert Steeves holds a Bachelor of Applied Science in Mathematics and Computer Science from the Simon Fraser University, and was a Software Engineer at Google from 2016 to 2018.
Contributors to Bittensor’s Academia and Whitepaper include Jacob Robert Steeves, Ala Shaabana, Yuqian Hui, François Luus, Sin Tai Liu, Jacqueline Dawn Tasker-Steeves, Yuma Rao, and the Opentensor Foundation.
NOT FINANCIAL ADVICE
Please note that Uphold and its affiliates do not provide investment, tax, or legal advice. This message is for informational purposes only and takes no account of particular personal or market circumstances, and should not be relied upon as investment, tax, or legal advice. For investment, tax, or legal advice and before taking any action you should consult your own advisors. Note that digital assets such as cryptocurrencies present unique risks for investors. Please see our disclaimer regarding risks specific to holding digital assets before investing.
Don’t invest unless you’re prepared to lose all the money you invest. This is a high-risk investment and you should not expect to be protected if something goes wrong. Take 2 minutes to learn more.
Share article
Share article
