Powering An Economic Internet With ML, AI, Multi-Agent Systems, And Decentralized Ledger Technology.
Share this article
Fetch.AI is a "decentralized digital representation of the real world in which autonomous software agents perform useful economic work". To do so, it relies heavily on Machine Learning (ML) and Artificial Intelligence (AI).
Fetch's goal is to "automate a limitless number of markets that require tremendous manual intervention". According to the team, autonomous economic transactions are expected to deliver near-frictionless marketplaces, operating at a very fast speed.
At its core, Fetch.AI relies on near-autonomous system integrating multiple components of a complex system, easy integration & use of ML / AI outcomes into decision making (without needing a thorough understanding of ML / AI techniques), and collective intelligence made of machine and human intelligence to create a non-siloed AI model, leading to improvement in decision-making processes.
The Fetch.AI (FET) token is used as the value exchange on the Fetch.AI network. It is the payment method for all transaction fees, and also a refundable method of registering on the ledger, and can be staked to secure the overall network.
1. What is Fetch.AI (FET)?
Fetch.AI brings together machine learning ("ML"), artificial intelligence ("AI"), multi-agent systems and decentralized ledger technology to create an economic internet — an environment where digital representatives of the economy's moving parts (such as data, hardware, services, people and infrastructure) can get useful work done through effective introductions and predictions.
Fetch's goal is to automate a limitless number of markets which currently require tremendous manual intervention. Autonomous economic transactions will deliver near-frictionless marketplaces, operating at digital speed.
Fetch.AI envisages a world in which everyone owns numerous economic agents on the Fetch.AI platform which operate in series or in parallel to construct solutions to today's and tomorrow's complex problems which are far more elegant and efficient than humans can produce.
Some of its key elements include:
Near-autonomous system integration for multiple components of a complex system.
Easy integration and use of ML/AI outcomes into decision making without too much understanding of ML/AI techniques.
Collective intelligence between machine intelligence and human intelligence to create a non siloed AI model, improving decision making processes.
Digital infrastructure for deployment of Multi Agent Systems.
Unique scalable ledger to support high transaction rates.
Synergetic computing: intelligent smart contracts for autonomous systems.
Economic framework for governing dynamic marketplaces.
Semantic, geographic and decision based navigation for autonomous agents to deliver seamless problem solving.
2. Products and metrics
Because the product is at testnet stage, there are no metrics on downloads, DAUs or any other usage or product metrics at this time. However, here's a conceptual diagram on how the Fetch.AI ecosystem is intended to work:
Consensus: Fetch.AI uses a combination of Proof-of-Stake ("PoS") coupled with a protocol to perform work to deliver consensus. New blocks are minted through the PoS protocol with the transaction order being determined by the work carried out between two blocks. This work is recorded on a directed acyclic graph (DAG) that grows between the blocks. The DAG is notarized by the blockchain which removes the need for a coordinator. A leader or committee of leaders is elected in each block epoch using a decentralized random beacon that is based on a threshold signature scheme.
Fees and incentives for participation : system incentives will be distributed to processing nodes for block production and for data mining operations on the smart ledger. Transaction fees will involve several components including a proposed gas price and a maximum number of operations.
The ledger is scalable by design, so its exact performance will vary depending on the configured resources available in the processing nodes. In private tests, Fetch.AI has achieved speeds in excess of 30,000 transactions per second. It is expected that Fetch.AI will gradually increase resources allocated to the test network as demand increases over time.
It is expected that at its inception, the test network will have a block time in the order of minutes. This conservative value will be reduced over time to reflect improvements to the consensus protocol and progressively more realistic test network deployments.
2.2 he Open Economic Framework
This is a second-layer protocol that provides services to agents. Agents connect to the OEF in order to find, negotiate with, and then do business with other agents on the network. It is designed to present agents with semantic, geographic and economic views on the world. Nodes can either be pure ledger nodes or ledger/OEF nodes. In the initial implementations, the OEF nodes come in two classes: "trusted" and "trustless". The "trustless" nodes can be operated anonymously, similar to the pure ledger nodes. The "trusted" nodes can be granted access to additional agent information to provide intelligent search and discovery capabilities. Operators of these nodes must have a public and legal identity, and receive accreditation by the Fetch.AI Foundation.
3. Economics and supply
3.1 Key metrics
ERC-20 (as of TGE)
Seed Sale Allocation
5.24% of Total Supply
Seed Sale Date
2017 to June 2018
Seed Sale Price
$$0.0341 USD / FET
Private Sale Allocation
6.38% of Total Supply
Private Token Sale Date
3rd April - 10th July 2018
Private Token Sale Price
$0.0680 USD / FET
$3,033,562 USD & 24,596.5 ETH
3.2 FET token ecosystem
There will be a fixed number of divisible tokens that are used on the Fetch.AI network as the digital currency for all transactions, as well as for network operations such as secure communications. Tokens can also constitute an access deposit for both nodes and agents wishing to perform certain operations (as a security mechanism to discourage malicious behavior).
Token allocation has been divided amongst public sale, seed investors & private sale, founders & team, advisors, ecosystem, mining rewards, and issuer.
3.3 Seed Investment & SAFT Raises
The issuer was Fetch.AI Foundation Pte Ltd.
Seed (2017 until April 2018): $2,058,587 Raised Amount raised fully in fiat% of Total Supply: 5.24%
SAFT 1: $5m round at the time of completion (3rd April - 30th April 2018). It raised $724,975 & 7,426.5 ETH % of Total Supply: 2.68%
SAFT 2 $10m round at the time of completion (1st May - 5th June 2018): Raised $250,000 & 14,448.99 ETH % of Total Supply: 3.17%
SAFT 3(6th June - 10th July 2018) Raised 2,721 ETH % of Total Supply: 0.53%
FET token distribution (%)
Token governance and use of funds
As of December 2018 1st, Fetch.AI has used approximately 85% of SAFT funds according to the allocations below:
5% Professional services.
All funds raised in the SAFT rounds were credited to a multi-sig account: two keys are held by TokenMarket (independent manager of private sale round) and two keys are held by Fetch.AI executives. Three keys are required to transact.
Liquidation of Crypto Holdings
Of the Ether raised across the three SAFT rounds (24,596.5 ETH), the majority (21,125 ETH i.e., 86%) has been converted into fiat currency during Q4 2018 and Q1 2019. As Fetch.AI costs predominantly originate from the UK in staff and office costs, this Ether has been converted to British Pounds.
For comprehension and convenience, the table below uses US Dollars.
Qty of ETH Converted
Weighted Average ETH Price
FET token release schedule
The following chart represents the number and breakdown of all FET tokens that are intended to be released into circulation on a monthly basis.
Note: Mining Rewards are intended to be issued in perpetuity at a decaying issuance rate.
Various tranches of token lockups and vesting schedules are detailed as follows:
Public Sale: no lockup after TGE.
Private & Seed Sale: 3-month lockup after TGE, vesting monthly over 3 months thereafter.
Founders & Team: 3-month lockup after TGE, vesting quarterly over 3 years thereafter.
Advisors: no lockup after TGE. Vesting quarterly over 3 years.
Future Release: targeted monthly release 12 months after TGE.
Mining Release: targeted monthly release 6 months after TGE.
Foundation: no lockup after TGE, vesting monthly over 4 years.
FET token release schedule
3.4 FET token overview and use-cases
When the Fetch.AI mainnet is launched, the Fetch.AI token allows for autonomous economic agents to get things done on the network.The Fetch.AI token is the key method of value exchange on the Fetch.AI network. It is required for all network exchanges, as a method of registering with the network, for staking and as a mechanism for delivering value back to those performing work on the network.
Fetch's token allows agents access to the digital world. It enables them to exist in this world, in multiple locations, and explore it looking for other agents to deliver value to or gain value from. This value can be in the form of access to services, data, infrastructure, or data processing via AI and ML algorithms.
Fetch.AI tokens can be used for many purposes, the largest six of which are:
Ability to connect agents and nodes to the network: FET is an access deposit token that acts as a form of stake to demonstrate desire to behave appropriately. It modulates the ability for bad actors to flood the network with undesirable nodes or agents due to the escalating cost of doing so. Fetch.AI are performing modelling and simulations on their testnets (both public and private) during the first half of 2019 to determine what this stake should be.
Value exchange between agents: The Fetch.AI token is required in order to allow for two agents, regardless of where they are, to perform a value exchange. The Fetch.AI token is infinitely divisible, thereby supporting transactions that have very low monetary value, but in aggregate provide new and profound level of insight and opportunity.
Access to search and discovery, and advertisement services: Fetch.AI tokens are needed to access, view and interact with the Fetch.AI search engine. This is a space optimized for autonomous digital entities, providing dynamic and autonomous price setting and negotiation.
Access to Fetch's multi-dimensional digital world: Fetch.AI tokens allow agents to view and explore its digital space geographically, semantically or economically. Some of these detailed views involve considerable computing time and communications between many nodes. These more resource intensive operations will require more tokens. Fetch's AI and ML constantly restructure this world to optimize each agent's individual view.
Ability to access and develop ledger-based AI/ML algorithms: The Fetch.AI token enables development of and access to a broad range of machine learning and artificial intelligence tasks that are available on the ledger. These may be Fetch-developed primary services such as trust and prediction models, or they may be large-scale independently developed services for network users.
For exchange into Fetch's operational fuel: Operation costs in Fetch.AI are decoupled from the Fetch.AI token in a similar way to that of "gas" on the Ethereum network, but with additional functionality designed to increase the stability of such a fuel and look at addressing issues associated with high and low-velocity economies. Fetch's operational fuel allows access to processor time for contract execution and services for agents.
3.5 Sample target use sases
Transport: Fetch.AI is working on reducing laborious travel planning tasks by using expert autonomous economic agents to organize complex trips, predict potential misconnections and dynamically reroute journeys, rearrange travel plans and rebooking reservations without intervention. The goal is to improve users' and transport providers' efficiency, service levels and satisfaction.
Energy: Fetch.AI is working on creating a fluid energy model to deliver the most effective energy solution to households without the friction of switching suppliers. Consumers could potentially change providers down to the minute or even by each appliance's unique energy demands. Fetch.AI is working with Warwick University on creating and deploying a live energy balancing simulation.
Supply Chain: Fetch.AI is enabling the trillion dollar steel sector to autonomously and collaboratively self-manage. It is now able to optimize its supply chain from the raw materials to the finished product, giving it the opportunity to massively reduce costs and improve efficiency. In addition, Fetch.AI is also working on building collective prediction models to improve efficiency in cargo rail route in collaboration with a German team.
4. Roadmap, updates, and business development
4.1 Communicated vs. completed milestones
When the TGE was delayed in Q4 2018 due to market conditions, Fetch.AI de-emphasized their public testnet and focused their efforts on the private network and building examples with the project's increasing number of commercial partners. Whilst Fetch.AI have incurred some delays in development, they have been minimal.
Released slightly late as part of the ledger code updates:Medium Link
4.2 Updated roadmap
Phase 1 - By Invite Testnet (Feb 2019)
Access to private test network extended by invite for network participation and wallets
Phase 2 - Public Testnet (Apr 2019)
Full public test network with smart contracts
Includes enhanced Open Economic Framework ("OEF") search capabilities
Phase 3 - Smart Market Testnet (Jun 2019)
Consensus, synergetic computing, & auctions
Full Fetch.AI consensus with multi-party & multi-dependency auctions and computing
Launch Phase begins (Jul 2019)
Delivers the initial mainnet OEF and ledger functionality in its initial form. All building blocks will be present
End of Sept 2019
All expected mainnet functionality undergoing tuning, optimisations and tweaks. Beta will extend out to the mainnet release
End of Dec 2019
Initial release of the Fetch.AI main network powered by the native on-chain Fetch.AI token (FET)
For Phases 1 and 2, the test network will employ a simple Proof-of-Work ("PoW") consensus mechanism. It is the intention that these phases of the testnet will allow developers to begin developing smart contracts and evaluating the Fetch.AI ecosystem of applications, libraries and services. In addition, the testnet in Phase 1 will be restricted to selected invitees — this primarily is to ensure a controlled ramp up of load on the network. Sign up will not be restricted but token holders will gain priority access to the network.
Fetch.AI expects that the node population will vary considerably based on public involvement. As such, Fetch.AI and some chosen partners will additionally run a number of core nodes in the network, and this is likely to be in the range 100-500 nodes.
In the third phase, up to June 2019, Fetch.AI will deploy a large update to the test network which will include the first release of Fetch's final consensus mechanism. This will be released alongside updates to the ledger internals and smart market support. The testnet will remain the primary network until the launch of the main network at the end of 2019.
At the beginning of the Launch phase, Fetch.AI will launch a bounty program to incentivize outside participants to find and report bugs in the source code. Additionally, Fetch.AI will start a development program to encourage the development of smart contracts and adoption of the Fetch.AI platform during the Launch phase.
4.3 Commercial partnership and business development progress
Blockchain for Europe : four leading blockchain companies including Ripple, NEM, Fetch.AI, and EMURGO/Cardano have come together to establish the "Blockchain for Europe" association. The association's goals are to foster the understanding within EU and member state institutions on the true potential of distributed ledger ("DLT") and blockchain technology, and to ensure that upcoming regulation promotes and boosts innovation in Europe.
Clustermarket: to further facilitate access to the right advanced technologies for the right science, GE Healthcare and Clustermarket have deployed Fetch.AI's Autonomous Economic Agents on the Clustermarket platform to provide personalized recommendations of specific equipment and services that may aid customers' research based on previous requests.
AI Innovation Network ("AIIN") / Warwick Business School: the AI Innovation Network provides a platform for collaboration across different stakeholders including public, private and third-sector organizations to examine the transformational impact of artificial intelligence. Fetch.AI is exploring various collaboration opportunities and discussing setting up pilot projects with the partners and members of this network.
ULedger: ULedger's stateless blockchain protocol makes it possible for the Fetch.AI network to consume centralized real-world data from IoT devices and other data sources in a provable manner. It brings security, integrity and identify to electronic communications.
MOBI: MOBI is a nonprofit organization working with forward thinking companies, governments, and NGOs to make mobility services more efficient, affordable, greener, safer, and less congested by promoting standards and accelerating adoption of blockchain, distributed ledger, and related technologies. Fetch.AI is exploring various collaboration opportunities and discussing setting up pilot projects with the partners and members of this consortium.
5. Project team
Co-Founder and CEO of Fetch.AI. One of the first investors of Google's DeepMind
Co-Founder and CTO of Fetch.AI. Producer of the a-life Creatures series of games and early developer at DeepMind
Co-Founder and CSO of Fetch.AI. Professor at Sheffield University and established scientist in advanced machine learning AI
Commercialization and Marketing
Head of Research
Head of Software Engineering
Head of Business Development
Senior Software Engineer
Senior Software Engineer
Software Developer & Ecosystem Coordinator
Senior Software Engineer
Senior Research Scientist
Lead Cryptography Engineer
Machine Learning Scientist
Machine Learning Engineer
Research Scientist Consultant
Senior Software Engineer
Machine Learning Engineer
Machine Learning Engineer
Tian Huey Teh
Machine Learning Engineer
Senior Software Engineer
6. Fetch.AI's activity and community overview
6.1 Development activity
6.2 Social and community data
Community growth is a tremendously important aspect of the Fetch.AI project. Fetch.AI has held numerous community-orientated meetups across the globe during Q3 & Q4 2018. This program will ramp up in 2019, particularly with the release of the project's testnet, leading up to mainnet launch. For example, these events could be augmented with hackathons.
Furthermore, Fetch.AI regularly showcases the work of different members from their team to explain the aspect of the project they are working on. This emphasizes the innovative technical challenges of the project. The response to these articles has engaged technical enthusiasts and opened up in-depth discussions in the blockchain development community.
During 2018, Fetch.AI also presented extensively at blockchain, AI and relevant commercial sector conferences across the globe. These platforms provide the opportunity to discuss the Fetch.AI vision and generates enthusiastic discussion and participation in the development community as well as interest in exploring commercial partnerships.
In Q1 2019, Fetch.AI will be releasing a dedicated developer community website where users can consume technical documentation, showcase their agents, discuss and collaborate on all aspects of Fetch.AI's platform and technology. It also allows for the co-ordination of meetups, hackathons, developer conferences and other activities where Fetch.AI and the community can interact with each other for mutual benefit. Fetch.AI values the developer community enormously: a vibrant, active and productive group of developers are essential for the growth of the network and the applications built upon it.
Participation and engagement in Fetch.AI's vision has been global in nature. It has been mostly active from independent developers who wish to build agents ready for mainnet launch and from corporate strategy and R&D teams who truly appreciate the application of Fetch.AI's platform to their sector(s).
In terms of social media growth, Fetch.AI has seen increased participation as a result of announcements, PR, and speaking at conferences. In addition, Fetch.AI will launch paid media campaigns in the run up to key dates in the project's calendar (e.g., token sale, testnet / mainnet launch).
Fetch.AI will continue to leverage wider communities - for example, sponsored video coverage and interviews with well-known influencers from the cryptocurrency, blockchain, AI and general technical fields.