Bittensor Paradigm
Written by Const, core dev @ Rao Foundation
Introduction
Bittensor's vision for a decentralized AI company has captivated a large number of people in the arenas of Artificial Intelligence and computer science, who have yearned for an alternative to the top-down world being created by our current technology giants. However the promise of what Bittensor is and can be has been masked by large technical barriers to understanding, and thus has not been well digested by non-technical audiences.
The purpose of this article is to answer why Bittensor matters to the world, why you should care and why we need to build it together. These are core questions, and each of them will require that we delve into the history of cryptocurrency, and how Bittensor is uniquely positioned to play a significant role in the future that is being written today.
Summary
The invention of Bitcoin was a revolutionary moment for humanity because it introduced two simultaneously united yet distinct technological innovations, reshaping the way that humans organize themselves. Primarily, it was the first ever decentralized currency, which gave humanity a shared value system - one that could not be mutated by vested or centralized interests. With this, Bitcoin promised a future of fair, anti-fiat, finance.
Secondly and to achieve this aim, it birthed the original, disseminated, and permissionless digital commodity market. This second point has not been overlooked by anyone who has been watching closely - the Bitcoin network's computational power, measured in raw hashing power, has blast past the potential of any company or government.
Bittensor is essentially a language for writing numerous decentralized commodity markets, or 'subnets', situated under a unified token system. These distinct markets function through Bittensor's blockchain, allowing each to interact and join into a singular computing infrastructure. By analogy, Bittensor brings the same type of abstraction which Ethereum added to Bitcoin for running decentralized contracts, but onto Bitcoin's inverse innovation — digital markets.
Compared to Bitcoin and other cryptocurrencies attempting to leverage the digital marketplace, Bittensor has built a framework that provides ease for creating these viable and enormously powerful systems. However, its genius lies in the fact that every one of these inter-networked markets is connectable, and available, to the whole. Building a hierarchical web of resources, ultimately culminating in the production of intelligence; intelligence leverages computation, which leverages data, which leverages storage, then finally leveraging oracles and data procurement and into infinity, all within the same ecosystem.
This is Bittensor's overarching vision; directing the power of digital markets towards society's most important digital commodity - Artificial Intelligence. Not only to build the most powerful intelligence network, but also to ensure that the benefits and the ownership of machine intelligence are in the hands of mere mortals. Bottom up, rather than top down.
Bittensor's goal and function of this technological foundation has several implications. For developers, a language to write markets for bespoke commodities such as compute, allowing them to take advantage of the enormous size, cost-efficiency and natural functionality of decentralized market systems. For front-end customers, Bittensor offers access to resources at a cheaper price and without intermediary;unclosable, not falsely self-proclaimed as “open”. For the Bittensor network as a whole, to create Machine Intelligence, in an open and equitable manner on top of ordered and dependent sub-markets. And finally - for the world, to ensure that the supremely important commodity of intelligence is owned by everyone.
Bittensor is, in a similar way, also a computer - albeit one with differing problems such as market misalignments - but one that is programmable, with the potential to become the most powerful ever created. And, it is a shared computer - not directly owned by any single entity, and available for anyone's use. To run, govern, contribute to, and importantly profit from the technological products it produces.
Understanding Bittensor
Understanding Bittensor is difficult for the exact same reason that it is powerful. This is because the majority of Bittensor's computational power is not concretely defined, but instead abstracted underneath markets. For instance, Bittensor leverages machine learning models to produce intelligence products, but Bittensor also leverages physical storage to create storage products. Meanwhile, it does not define the specifics of how these resources are produced. Instead, Bittensor simply defines the market which will reward those commodities for becoming available to the network.
Abstractions like this are the crucial element to how Bittensor is able to leverage enormous swaths of resources: without defining the way in which a computer interacts with Bittensor's markets, the market leaves space for innovation around how that problem will be solved. In other words, without directly purchasing or inventing the computers that plug themselves into the network, that problem is left to thousands of incredibly talented individuals all across the Earth, who achieve that aim in various and creative ways.
Bittensor vs Bitcoin
As a good starting point for understanding how the rubber hits the road, we can start by comparing Bittensor to Bitcoin. We use Bitcoin as it is the ultimate example of digital markets in action, and has the benefit of being well understood by many already.
Bitcoin is secured by a game theoretic equilibrium, whereby it is much more profitable for miners to accurately and truthfully replicate the state of Bitcoin's account balances, rather than diverge or create false histories. Bitcoin achieves this equilibrium by distributing rewards to its network in rough proportion to the quantity of compute a network participant brings to the network, thus creating a market for compute power.
Bitcoin itself, i.e. the code of the protocol which stitches this equilibrium together, does not itself run any computers, nor in recent years does it even define the way in which computing power proves its existence to the network. Rather, it defines the way in which computing power is validated - that is, the manner in which the computers in the system agree together who has the most computing power and thus the most reward in the system. This is a digital commodity incentive mechanism.
The effects of this abstraction have dazzled many. Over the last 12 years, Bitcoin's provable computing power has exceeded 1000x times the equivalent potential of the trillion dollar company Google, and is set to consume more electricity than the entire nation of Australia. But how did Bitcoin achieve this?
In summary, digital commodity incentive mechanisms are the perfect markets, and perfect markets have the amazing quality that, when aligned, they are unstoppable and unequivocally powerful. For instance, Bitcoin miners worldwide continuously and aggressively find ways to reduce the cost to mine Bitcoin by purchasing cheap hardware, or finding unused lakes of available energy. The system is antifragile, constantly killing off bad competition. It is unclosable. Anyone can join, no questions asked, and contributions get paid directly and continuously, no contracts. No HR department − this is capitalism in its purest form.
Bittensor is like Bitcoin in many ways. It has a transferrable and censorship resistant token, TAO, which runs on a 24/7 decentralized blockchain substrate which is auditable and transparent. Bittensor is also run by miners, like Bitcoin, who can exist globally and anonymously.
Importantly and conversely, Bittensor utilizes these same advantages of Bitcoin, which has incentivized its compute network to grow enormously. For instance, when a market is defined on Bittensor, miners worldwide continuously and aggressively find ways at reducing the cost to mine TAO on that market; they purchase hardware, stay up all night, and find every way to reduce costs. Procuring all available resources whether from their own home, or from datacenters worldwide, both known and unknown - all to create the digital product of intelligence, or anything and everything the network defines. In essence, we leverage the same killer app.
Where Bittensor differs primarily is the product. Whilst Bitcoin focuses exclusively on the importance of ensuring Bitcoin's total resistance to exterior actors and the immutability of its token economic system, Bittensor focuses primarily on building value-creating markets.
Focusing here; the Bittensor paradigm is about being a network, computer and language for writing systems, alongside the market which created Bitcoin's gargantuan size. However, they fundamentally differ in how they use these resources. Where Bitcoin simply secures its network by creating one trivial product (namely SHA-256 Hashes - which are unusable and not readily transferable into any real problem), Bittensor doubles down, allowing the building of multiple intertwined and complementary sub incentive systems, which build out commodities, like data or bandwidth, with real-world, applicable value, culminating ultimately, in intelligence.
Bittensor's Core Tech
Arguably, Bittensor's core technological innovation is the separation of the chain's core functioning (transferring funds etc.) from the running of the validation systems which define the markets for its digital commodity creation. This is distinct from Bitcoin, Ethereum, Filecoin and most every digital commodity system on the market today
Because of this separation, Bittensor does not need to have a special purpose language for writing these validation systems. They can be written in Rust, Python, C++, or any other programming language. Furthermore, all of the tools required for validating the subnet mechanisms remain off-chain, allowing them to be potentially extremely data heavy and compute intensive.
For instance, in Bittensor's flagship commodity market, Subnet 1, intelligence providers are validated for answering queries using a machine learning model. Those outputs are themselves validated using machine learning models, expensive and large ones, which can subsequently learn and be adapted without any information moving between the validator set and the chain. This would be impossible on nearly any available blockchain.
What makes this possible is Yuma Consensus (YC). Yuma Consensus is the mechanism Bittensor runs on its chain, to enforce that there is agreement between the validators of the individual sub mechanisms in Bittensor's network. By analogy, Yuma Consensus is the central processing unit (CPU) of Bittensor. YC takes the varying incentive mechanisms written by developers and transforms them into an incentive landscape, whereby agreement is reached and miners are forced to act inthe way defined by the people who write the consensus mechanisms for the individual subnets.
The unique innovation of YC is that it is agnostic to what is being measured, and allows for fuzzy consensus around probabilistic truths - truths like intelligence. Indeed, YC was essential for the validation of intelligence, given the nature of that amorphous information theoretic commodity.
YC is what makes Bittensor fundamentally distinct from various other blockchain projects - most of which write their validation systems directly into their chain and thus make the problem being solved, or the way it is solved, immutable down the road. Yuma Consensus has no such limitations.
Why should business people care?
The ability to write a multiplicity of Bittensor incentive systems and have them concurrently run within a single token ecosystem is not just cool, it's absolutely essential for building decentralized organizations which can compete with large corporations. Google has an AI team, backed by a storage team, plus a compute team all under one roof - whilst you (without Bittensor) are forced to use 10 different token economic markets on 10 different chains.
Bittensor gives you all of it: an AI team, a storage team, a compute team, and anything else that can be dreamed of - all under a single token framework, turning TAO holders into effective Decentralized Cloud Providers, backed by an incentivized, adapting and ever-improving infrastructure. TAO is the bridge between clients and this mega-computer, plus the resources that are being created, through various subnetwork mechanisms, in Bittensor. This is the main functionality of TAO. Functionality, which hitherto was only within the reach of immense supercorps like OpenAI, is now accessible - in one non-permissioned place.
Bittensor's technology was built for developers primarily. But, just like any Information Technology, Bittensor enables business opportunities. Whilst its initial support is vastly driven by a community of technologists, Bittensor's impact will multiply once business professionals also understand the potential of decentralizing applications powered by intelligence
What's in it for developers?
The exercise is now to find old applications that are slow, expensive, archaic, highly-permissioned or not meeting the required needs, and figure out if they could be re-thought via incentive mechanisms, for the same things. Developers must learn about digital commodity markets, and ask if old processes can be replaced by open ones.
If you fall in one of these categories, you are in an excellent position to start diving into Bittensor:
- Machine learning engineers
- Intelligence companies
- Trading firms
- Network infrastructure companies
- Data acquisition companies
- Oracles
- Infrastructure providers
- Mining teams for all of the above
- Anyone who wants to have a say in the future of technology.
Bittensor is open to innovators and, for the first time, developers have the opportunity to monetize their ideas for the grand resource allocation system that they want. Drawing together huge quantities of resources, under incentive driven compute systems -without the need to engage in the creation of an entirely new chain. Bittensor provides a platform to build these systems in one place. It thus provides a one-stop-shop for those seeking all the compute requirements for building unstoppable applications on top of an incentivized infrastructure.
Open Markets, Open Ownership
At its core, Bittensor is a horizontally scalable, cryptography-based, resource-allocation protocol that is also open to contribution. However, what makes it so powerful from a technical standpoint is what makes it equally attractive from a humanistic perspective. That's because the inverse of Bittensor's ability to attract resources from the wider world, is its ability to distribute real visceral ownership back out to that same world.
In essence, Bittensor's permeability to contribution is its permeability to control. You can own the network in the same manner I, as a long-term developer of Bittensor, can own it. This is a quality of open ownership systems which, like its predecessor open source software, reevaluates what makes things good.
In open source we give the code away; in open ownership we give the control away, to the same effect and from the same principle - that which is the most open to contribution and control is the most open to innovation, and therefore most likely to align with the largest group of people.
This is as simultaneously vital - the only thing capable of competing with the largest corporations is the unity of everyone - as it is ethical. The excessive centralization of control, especially around AI, poses the greatest risk to the human race. The concentration of power inevitably creates biased decision-making, controlled access to benefits, and significant abuse.
Satoshi understood that the centralization of power over money inevitably creates corruption. This is the reason that Bitcoin sought to decentralize finance. This truth maps self-evidently onto humankind's most valuable digital resources. AI must not be owned by the few for the same, and exponentially more dangerous, threat.
Conclusion
It is important to revisit that there is often a parallel between what should be for technological reasons, and what should be for bigger issues facing the human race.
Technologically, Bittensor makes sense; abstracted computing markets are the most powerful tool humanity has ever had for aligning digital resources.
And humanistically, open-ownership systems are tools for the bottom-up placing of these resources, directly into people's hands. Bittensor is the ecosystem and TAO is the mycelial network connecting humanity with the future of Artificial Intelligence, such that its control can be ethically united with a life-affirming and truly democratic future for humanity, alongside machines.
We are approaching a forking point for mankind; down one road is the centralization of power and resources, in large regulated industries, who have entrenched and overbearing access to the best-in-class intelligence and computers. Down the other road is the potential for sharing these resources through open protocols, via technological foundations, which enable global participation and ownership.
Bittensor finds itself at the precipice of a completely new-age of computing: supercomputers via abstracted markets. This precedence shines forth by taking the greatest financial innovation our planet has ever seen, then wedding it to the real world: Bitcoin meets Artificial Intelligence. We can do it. We have the people, we have the desire, and we have the technology.