SingularityNET: The Case for a Trillion-Dollar Protocol
The world is shaping itself into something that would have been unrecognizable just ten years ago. With astounding innovations like tactile augmented reality, self driving cars, bio-engineered neural computer chips for our brains, and much more, it is a trustworthy statement to say that the next ten years will look nothing like the last thirty.
At the center of all this innovation, it is commonly believed that the most important technology humankind will ever create is artificial general intelligence.
This belief is a powerful concept immortalized in the popular non-fiction novel, Our Final Invention: Artificial Intelligence and the End of the Human Era, written by James Barrat.
The book lays out the potential benefits and possible risks of a human-level or even super-human artificial intelligence. It develops a framework which allows the reader to grasp the notions of what a future could look like with the advent of the Singularity: the moment artificial intelligence becomes super-intelligent.
The time it will take for us to reach that moment is unknown, yet it is within that window of time that AI will vastly reshape the way the world works. In a world empowered by AI, SingularityNET will not only be the best performing asset of the decade, but also the most important and beneficial technological ecosystem of our lifetime.
What is SingularityNET?
SingularityNET is a decentralized marketplace for AI algorithms that is secured by blockchain technology. It is the first and only platform that allows AI to cooperate and coordinate with each other at scale. The SingularityNET platform isn't just a commercial launchpad for AI developers to build AI services across any sector, or a service for client-facing applications — it is built also to nurture a decentralized “artificial cognitive organism.”
Anyone can launch an AI agent (node) on the SingularityNET platform in which the agent can autonomously interact with other agents for beneficial economic needs. This will allow agents to cohesively work together while being completely independent with their own unique prerogatives.
The dynamic here is if an AI agent needs access to data that it does not have, it can communicate with other agents on the platform to source specific data to utilize within its own operations autonomously. This means the platform can provide custom AI tooling to end users by contracting out a single AI Agent at a fraction of what they would pay large tech companies for.
This is the framework which will nurture a cognitive synergy. Why? Because while narrow and specific AI interact with each other to fulfill their operation, higher level organizational algorithms analyze those data transactions and learn to recognize patterns, which in turn allow those algorithms to learn how to learn.
This organizational formulation of patterns will not only allow the platform to run more efficiently through cognitive synergy, but also be the first artificial reflection of our own human neural networks. These meta-learning algorithms and concepts are still being researched and stress-tested by SingularityNET and OpenCog – a research project for open sourced artificial general intelligence software.
So, while SingularityNET provides a decentralized and globally distributed open market for AI tooling and products, it is concurrently laying the initial foundation for the first artificial ‘global brain’.
But before we can understand just how remarkably well-positioned SingularityNET is on a global scale, we need to take some time to understand what's under the hood.
Under The Hood
SingularityNET is a protocol of smart contracts that aims to be blockchain-agnostic, but is also porting from Ethereum to Cardano to begin its foundation. At its core, Agents – which can either be autonomous nodes on the network or external software that governs an AI Agent – utilize smart contract templates to request AI work be done, exchange data, and supply results from work done.
There are four main architecture objectives SingularityNET aims to accomplish:
- Creating a protocol that can be interoperable with any blockchain.
- Speeding up AI development on the platform.
- Creating a marketplace that is regulation-compliant.
- Laying the most robust foundation for beneficial artificial general intelligence.
The AI Marketplace
The AI Marketplace is the cornerstone of the SingularityNET platform that is underpinned by complex smart contracts and design implementations like the Registry, the Multi-Party Escrow, the Daemon, and the SingularityNET SDK.
The Registry is a smart contract that allows the SingularityNET platform to stay compliant by forcing the AI developer to describe their AI service for market participants to analyze.
The Multi-Party Escrow is also a smart contract that handles payments through escrow accounts using atomic unidirectional channels for cheaper and faster transactions. The Daemon is the off-chain side-car proxy that checks whether all escrow conditions are being met, and then validates requests with the AI service. This allows AI services to make money as their product is being utilized. Lastly, the SingularityNET SDK is a compiler for client-side code so that application developers may use SingularityNET's network intelligence.
The reputations system that SingularityNET implements is more than just an incentive structure for intra-platform activity, it is also the underlying framework for their governance and economic model. In the future, the reputation system can also be the consensus mechanism for SingularityNET that will allow the platform to be multi-chain.
Both AI services and Agents will be subject to the reputation system, yet both will likely be rated by different factors. For AI services, machine learning algorithms can analyze various types of data like:
- Financial Transaction Trails
- Explicit Consumer Ratings
- Past Ratings (time-decayed and/or weighted)
- Escrow Patterns (arbitration triggers, early cancellation etc.)
This will allow AI to adequately differentiate between malicious and non-malicious services, highly performant and non-functional services, and custom choose nuanced services based off of prior consumer needs. This incentivizes AI developers to create highly-efficient and beneficial products.
As for Agents, a base reputation system from 0-5 is implemented to justify the types of roles an Agent has in the ecosystem. An Agent with a rating at or above 2 can participate in on-chain governance. SingularityNET is still developing a scalable and efficient level of consensus for Agents, researching things like liquid rank, reputation policing, and domain-specific reputation.
As for the economic model, Agents (nodes) are incentivized to hold good reputation so that they can attract more stake from AGI token holders, and also participate in on-chain economic activities. Each Agent, whether AI or human, must have an operator that is KYC compliant.
For now, SingularityNET uses centralized KYC companies to do this, but in the future may implement a federation of AI Agents specifically for KYC'ing operators. The combination of amount staked and reputation is factored into the weight of an Agent's vote. This incentivizes operators to maintain healthy relationships with end users by providing only high-rated services.
The goal for SingularityNET to be maximally interoperable is contrived from a well known problem within the cryptocurrency industry: Ethereum gas fees.
When SingularityNET first launched in 2017, the only smart contract platform that existed was Ethereum. Running on the built-in non-scalability of Ethereum, a future in which SingularityNET can run trillions of API calls (API-of-API's) was not attainable. Thus, they made the decision to port to Cardano.
Cardano is an emerging generation-three blockchain that is highly regarded for its scientific approach to building blockchain technology.
The Plutus Integration
In regards to being highly scalable, SingularityNET also chose to port to Cardano to utilize their smart contract language, Plutus. Unlike Solidity – Ethereum's smart contract language – Plutus (based off of Haskell) allows for flexible capability for abstraction, which works well with creating an API-of-API's.
Also, using Plutus, there is a possibility to embed their API structure into the smart contracts themselves. On Ethereum, the only way to implement this would be to create a separate software layer above the smart contract layer, which smart contracts would need to invoke.
Cardano also built a DSL (domain specific language) called Marlowe for finance that is very similar to what SingularityNET is building for their AI Agents. SingularityNET is planning to use the scripting language Idris for their AI-DSL and build an Idris-to-Plutus compiler.
Again, the main reason why Cardano is so beneficial for their AI-DSL is because Solidity would force SingularityNET to build a separate level above smart contracts for smart contracts to invoke on Ethereum. Plutus can allow the AI-DSL to be written in the smart contracts themselves.
The Cardano-SingularityNET integration may even play an important role in decentralizing the way that the SingularityNET platform aims to create the collective society of AI minds like we talked about above.
OpenCog Hyperon Integration
Using Plutus it may be possible for OpenCog Hyperon – an initiative from OpenCog that will scale the current OpenCog platform – to be provided as a service to individual AI Agents who would need to utilize it. Currently an AI Agent would have to encapsulate the entire OpenCog Hyperon design upon creation every time an AI Agent is deployed.
OpenCog Hyperon will provide the fabric of which can deploy cognitive processes and learning/reasoning methods like probabilistic logic, evolutionary learning, and pattern mining to the SingularityNET platform.
On top of this, Plutus may also provide the AGI token (SingularityNET's native token asset) additional governance-related smart contract features which will enable greater security measures.
With Cardano, SingularityNET has a much better chance at building a multi-chain future for itself. Soon smart contracts from Avalanche, Ethereum, and Cosmos will be able invoke AI services on the SingularityNET platform. Stay tuned for upcoming Cardano 360 events to be updated on different implementations of the SingulrityNET/Cardano integration.
Now that we know what's under the hood, we can talk about the global impact SingularityNET can have on the world.
The Fourth Industrial Revolution
The First Industrial Revolution was catalyzed by steam power, the second by electric, the third by digital, and the fourth will undoubtedly be by the power of AI.
Ever since the first industrial revolution, humankind has become exponentially more advanced. The velocity, scope, and magnitude at which technologies will affect the way we operate is about to hit terminal velocity, and it is all culminating within the decade.
The reason AI is the driver of this revolution is because it is the benefactor of all other parallel technological advances. Robotics will use AI neural networks, genetic engineers will use data analyzed and aggregated by AI, AI will intertwine within the Internet of Things, AI will empower decentralized finance and blockchain — I could go on.
The global impact AI will have on the global economy is documented very well by PriceWaterhouseCoopers – the worlds second largest accounting network. Their projections illustrate a global AI economy worth $16 trillion dollars and being the direct cause for a 14% global GDP hike by 2030. To put that into perspective, the United States GDP is roughly $20 trillion dollars. Projections include a 55% increase in labour productivity, 27% in quality production, and 19% in time saved.
There are two main regions we need to talk about:
- North America
North America is projected to account for $3.7 trillion dollars of global GDP from the AI economy.
The reason this is important is because of the way AI is being developed in these regions and how, over time, they will affect the way we view AI. Each region poses their own unique opportunity for SingularityNET to capitalize on.
North America is the world's most AI-ready country, and is set to capture the most amount of initial gains made in the artificial intelligence revolution. The three main sectors that will benefit most for North America is:
The three major areas in which AI can benefit is personalized financial planning, fraud detection, anti-money laundering, and process automation.
The reality is that, while this may be true, an entirely new revolution is taking place within blockchain called decentralized finance (DeFi). The prerogative of DeFi is to provide a trustless and more inclusive alternative to traditional finance. There is already decentralized borrow/lending markets, decentralized exchanges, and even decentralized insurance. The valuation for something like this is unknown but could very well capture a majority share of the legacy financial installments.
So while there seem to be two separate financial worlds out there right now, SingularityNET is poised to capture both. The SingularityNET Foundation is developing two separate products that will run on the SingularityNET platform: one to capture value in the DeFi world, and one to capture value in the TradFi (traditional finance) world.
The former is a project called SingularityDAO. This project will create ETF-like baskets of low-liquidity utility tokens dynamically adjusted by AI algorithms. You can earn a yield by providing liquidity to the protocol and also utilize futures-based hedging features. AI-managed portfolios in the TradFi world outperformed the S&P 500 by 72%.
The latter is a a project called Singularity Studio which is actually an enterprise solutions AI software that runs on the SingularityNET platforms that covers industries like fintech, healthcare, IoT, and manufacturing. For fintech specifically, AI services can automate KYC/AML, estimate hedging risk, detect fraud, rate consumers on credit for the banked and the unbanked, and also analyze regulatory compliance for fintech companies.
The United States healthcare market was worth $9 trillion dollars last year strengthened by a decline in oil prices and a surge in elderly population. There are vast opportunities in the healthcare market for AI. The three largest are micro-pattern recognition for detecting diagnosis, potential pandemic tracking, and imaging diagnostics. Here are three ways SingularityNET can capture them:
The first is the project we listed above, Singularity Studio, which provides the following healthcare AI services:
- Planning and Resource Allocation
- Automated Personal Assistants
- Automated Image Diagnosis
- Claims Management (Insurance)
- Fraud Detection
- Big Data Analytics
Singularity Studio will be a product on the SingularityNET platform that individuals, corporations, and enterprises can utilize.
The second way is with SingularityNET's deep-rooted partnership with Hanson Robotics (known for the creation of Sophia, the Robot).
Dr. David Hanson Jr., the founder of Hanson Robotics, is also the co-founder of SingularityNET. Sophia the Robot's neural network is run using the SingularityNET SDK. Together SingularityNET and Hanson Robotics created Awakening Health.
Awakening Health is a combination of AI, biotech, and robotics. Hanson Robotics places most of their robotic design on the importance of facial expression. Facial expression is the most important non-language social cue that humans use for social interaction. The creation of Grace, the medical assistant robot and sister of Sophia, will mark the beginning of a new era of robotic AI medicine.
The third way is with a project being built on the SingularityNET platform through the SingularityNET Foundation called Rejuve.
The heart of Rejuve's mission is to combine AI, blockchain, and biotech to create therapies that will increase the human lifespan in ways we can't yet imagine. The Rejuve Network will allow individuals and clinics to control their own biodata and be compensated for the rights to use that data for specific therapies, pharmaceuticals, and gene therapy.
Rejuve will be able to collect user data from smart devices (like Apple Watches) and store them in a data-sovereign way, unique to the user so only the user can issue rights to that data.
Rejuve also has the rights to develop solutions using the data from the Rejuve Network. The problem is that genomic biodata production increases exponentially without sufficient ways of analyzing that data. Rejuve aims to leverage SingularityNET AI to sufficiently analyze big data sets in ways that humans physically cannot.
Nearly 12% of the United States economic output ($2.4 trillion dollars) and 9% of the total employment is thanks to the manufacturing industry. Manufacturing may be the most transformative industry (in both good and bad ways) with the advent of AI and blockchain.
SingularityNET is a lightweight network that can run on computer chips or embedded devices like robots in manufacturing and other IoT devices. This will allow embedded devices to utilize the SingularityNET platform and neural network and acquire cognitive synergies. This will boost efficiency in manufacturing robots, allow for self-upgrading procedures, and identify detrimental factors in the manufacturing system.
The African AI economy is projected to account for $1.2 trillion dollars of global GDP by 2030. Africa poses one of the most unique opportunities for technological exposure, and has the most vertical potential to leapfrog existing developed countries. Africa is human-resource rich, which means it is data rich — and in a world where the most valuable asset is data, this puts Africa in a unique scenario.
Africa has been remarkably open to embracing new technology, as they have no long-lasting trust in their own legacy installments. While this opportunity does exist, there needs to be more inclusive ways in which technology can pierce through in order for this leapfrogging to take place.
Africa is arguably least positioned out of the worlds regions to take advantage of the artificial intelligence revolution. This means there are different sectors that would see the most impact in creating a more sustainable and efficient Africa. The three main sectors that Africa would benefit from AI in the long term is:
- Internet-of-Things (IoT)
Sub-Saharan Africa is a largely undeveloped part of Africa without sufficient data to project accurate analytics for future education markets. With that being said, education is one of the most important sectors to grow in a developing economy. It is the fastest way to cull information asymmetry, and opens up job opportunities across various sectors.
The combination of Africa's willingness to embrace peer-to-peer technology, and the growing trend of decentralized education, Africa has a unique opportunity to take advantage of personalized education, e-learning, and automation curriculum.
Africa has also been the fastest adopter of handheld devices like mobile phones and tablets. This provides a direct bridge between end-users and machine-learning, algorithmically-designed, educational curriculum personalized for end users.
SingularityNET partnered with UNESCO's International Bureau of Education to develop a new form of curriculum spanning the years from kindergarten to the 12th grade. They are rolling out plans to span across six different countries to teach an extensive array of subjects such as robotics, natural language AI, machine learning etc. This path towards global digital inclusion with help SingularityNET foster IoT deployment in Africa as well, which lends its hand to its other projects on the platform.
2. Internet-of-Things (IoT)
The Internet-of-Things is defined by any interrelated computing device with a unique identifier that can transfer data over a network. As we mentioned above, the convergence of peer-to-peer transfer, mobile broadband penetration, and Africa's eagerness to capitalize on emerging technology, again positions itself well to capture that value.
The existing roadblocks though (such as the lack of electrical grids, expensive internet connection, and insufficient infrastructure) pose formidable obstacles.
Yet, as those roadblocks become less formidable, SingularityNET becomes exponentially more beneficial for African IoT infrastructure. SingularityNET is also partnered closely with another project called NuNet. NuNet is a computing platform solution that is capable of optimally providing globally distributed computing power and storage for decentralized networks.
NuNet will connect owners of data and computing resources with computational processes that demand them. NuNet will allow SingularityNET Agents to access the network of computing resources to further optimize their resource capacities. NuNet will allow users to harness computing resources in a globally distributed network. This will in turn provide further data inclusiveness for other SingularityNET Agents to utilize.
Singularity Studio is again another enterprise-level tool that will help IoT infrastructure in Africa. Here are some ways Singularity Studio can help:
- Collaborative and Context-Aware Robots
- Predictive Maintenance
- AI-Enhanced Supply Chain Management
- Yield Enhancement in Manufacturing
Manufacturing is a profit-first industry and historically has done everything in its power to increase profits; this has dark connotations behind it, as most profit-first operations do.
This also means that while regions like rural China, east Asia, India, and rural Europe benefited from developed countries leveraging their economy for cheaper labor by being able to provide jobs, Africa might not get its chance. This is because developed countries are outsourcing redundant labor jobs to robots, 3-d printers, and computerized stitching instead of contracting cheap labor.
The educational arm of SingularityNET may change the way that Africa can take advantage of this sidestepping of global manufacturing by changing the curriculum of eduction to be able to program and maintain the automated workers of the developed world via SingularityNET's platform.
Cardano and SingularityNET in Africa
I wrote an article covering the current financial and agricultural landscape in Ethiopia and how Cardano can help revolutionize the way Ethiopia conducts business using blockchain.
Cardano just partnered with the Ministry of Education in Ethiopia to onboard nearly five million students and one million teachers. They also partnered with World Mobile in Tanzania to onboard 100 thousand mobile phone users onto the Cardano platform.
SingularityNET has been in Ethiopia since 2014 working through iCog Labs, a partner of SingularityNET, to foster AI development and education for Ethiopian students. Cardan and Africa are planning on expanding efforts to foster blockchain and AI development. This will help commercialize the SingularityNET platform much faster by leveraging AI developers in Africa instead of competing with big tech companies for with deep pockets for AI developers in the West.
The Decentralized Perspective
There is well-placed fear in the idea that AI will foster the most monopolistic environments the world will ever see. The Big Tech companies like Apple, Google, Facebook, Microsoft, Amazon, and IBM have amassed multi-trillion dollar valuations over the past few years.
They aren't the only ones either, Chinese tech giants like Baidu, Alibaba, and Tencent as well. Most of these companies have a previous history with anti-trust laws, mainly Facebook. Each one of these companies is the most well-positioned to source AI developers, foster multi-national AI deals, and spend obscene amounts of capital on AI R&D and M&A.
When the world inevitably becomes more artificially intelligent, these big tech companies will begin siphoning market share from every sector of the economy because they will be the ones with the best products and services on the market.
This silo-ing of AI development will have reverberating impact globally, likely in more ways than you think:
- Firstly, these multi-national companies will force emerging market countries (who desperately need AI) to purchase their products at a premium and eat into their national budget. This will siphon money from every sector of the globe as well.
- Secondly, this will create nation-state insecurity as these companies will become more wealthy than entire countries.
- Lastly, and most importantly, because of the positive feedback loop of data-fed AI and big tech companies access to enormous amounts of data, their AI will grow exponentially relative to its counterparts.
This means it becomes ever more likely that big tech companies will be the fore-fathers of the first artificial general intelligence. The reason why we don't want this is because AI will form inherent biases from the environment around it. Like humans, this transitory phase from narrow to general intelligence will be encapsulated by two pillars: environment and evolution.
The AI will be subject to the data and algorithms that once defined it, and those algorithms and data feeds will have undoubtedly been written and provided in a profit-maximizing way. This means the basis of the first general intelligence will form its biases based on the foundation of efficiency over benevolence.
This also means that underdeveloped countries that don't have any exposure to such an AI will be discriminated against by lack of AI inclusion.
SingularityNET provides a globally distributed, decentralized, and mutually-benefitting platform which can provide much cheaper alternatives to just as, or even higher quality, AI services and products than its big tech counterparts.
The platform, like we have talked about, was built to promote the emergence of the first AGI. In an ecosystem in which developers are economically incentivized to maintain quality reputation by providing beneficial AI services, and operators of Agents are economically incentivized to to maintain quality reputation by brokering consumers with helpful products, the underlying meta-learning algorithms can nurture a truly benevolent AGI.