Today we’re going to continue our look at the Binance Launchpad IEOs with Fetch.AI. This project is one that I quite frankly didn’t understand until I really decided to delve into the details of how it would it function. With this article I’ll be going through what Fetch AI is, the three major components of the project, and a brief outlook on its roadmap progress. As usual this is an overview of the project and is not meant to be comprehensive.
Fetch.AI is, according to the technical introduction, “…a decentralized digital representation of the world in which autonomous software agents perform useful economic work.”
What that means is Fetch.AI is constructing a network where “Autonomous Agents” that represent people, businesses, services, are all able to interact with each other and exchange goods (information) that provides economic value. In order to accomplish that task Fetch is creating three layers of architecture the Autonomous Economic Agents, the Open Economic Framework, and the Smart Ledger.
With those components the digital economic representation will be able to be used to solve issues of inefficiencies in large scale industries like shipping and energy as well as performing more simple tasks like booking a hotel or a plane ticket. Now let’s dive into the components of Fetch.AI.
The AEA are the workhorses of the Fetch.AI network and these are independent programs that perform some type of task. Taking a step back though, the Fetch network itself is made up of a network of nodes that help to act as a directory for the agents and for the record any install of Fetch is considered a node by the network.
An agent then can be anything connected to the network including low power devices like a sensor. Now being an agent means many things including being able to be programmed to interact with other agents through the sharing or purchasing of information. The agent can also be used execute a command without human input. Furthermore, these agents have a memory that can provide which other nodes they interacted with and why.
To learn more about agents please see this blog post from Fetch.AI.
After the AEA we have the OEF. The OEF is the value exchange framework for how agents can interact with each other to give or even purchase requisite information. What that means practically speaking is the AEAs live within the OEF and that framework permits agents to search and interact with other agents.
The OEF itself is “a combination of APIs, directories of services and agents, previous transactions, wallets and agent positions” which can also be described as a ton of information. This information is open to be being queried by the agents and the OEF then seeks to find the information and then place the agent seeking the information to the node closest to the agent that can provide the information.
That process can take some time, but once the process is complete the information regarding that query and result is broadcast to the entire ledger and OEF. What that means is there will be memory of that search and next time the information is required the search and result can be performed much more quickly. Lastly, with the OEF “placing” agents closer to nodes that provide them useful information for their tasks the agents themselves will eventually be placed in an optimal position for them to perform their tasks and provide other information making this framework fluid and adaptable.
To learn more about the OEF please see this blog post from Fetch.AI.
The final layer of Fetch is the Smart Ledger which aims to provide Fetch with scalability among other things. The ledger itself is a partly a directed acyclic graph or DAG. The structure of the DAG is that of an ever expanding tree as opposed to the “link” like structure of Bitcoin. For a complete rundown of DAG’s vs Blockchain take a look at this article.
Ok, so Fetch uses a DAG what else? Well, it combines the DAG with sharding (in a sense) and will also be able to be permisionless or permissioned depending on the needs of the user. Now, you may be asking why Fetch needs this big fancy DAG as opposed to a blockchain. The reason for that is due to the massive amount of information that will need to moving around at any given time if growth to full adoption occurs. The Fetch system could be used to query and find solutions for supply chain issues, energy problems, health information for populations, and other issues all potentially at the same time.
In the DAG there are subnetworks that are created to help process information faster. Fetch compares the creation of additional subnetworks to the addition of side chains to a blockchain or a shard. Fetch feels their solution here is superior to that of the side chain since there aren’t any security or performance bottlenecks that may occur thanks to the subnetwork design which they refer to as “resource lanes.”
For a full breakdown of the innovations that Fetch has put into the ledger to prevent double spending and optimize for speed and security check out this article from Fetch.
The Fetch.AI roadmap is comprised of four sections for 2019 broken down by quarter. Currently the big goal is to release mainnet by the end of 2019. The biggest releases that have been put out are the public testnest as well as a block explorer and smart contract playground. To remain up to date on the roadmap and Fetch’s progress I’ve found that the best resource for major updates is this page on the Fetch blog.
To learn more about the Fetch.AI project please see the following links:
Website – https://fetch.ai/
Whitepaper – https://fetch.ai/uploads/technical-introduction.pdf
Fetch FAQ – https://community.fetch.ai/faq/
If you would like to get started developing with the Fetch.AI VM or learn how to code your own node check out the Fetch.AI community website
I want to make it clear that with this article I have only scratched the surface of the components of Fetch.AI that make it a compelling and interesting topic. I suggest that those who are interested read the technical whitepaper that Fetch.AI has published. Furthermore, the technical qualifications of the team are quite impressive and the team already has a working public testnet that was released almost exactly one month ago on April 30th 2019.
Over the past half year I have grown quite weary to the idea of AI and Machine Learning crypto projects as long term they haven’t seemed to really pan out. That being said I think that Fetch.AI has the components that may allow it to buck the trend.
Thank you for reading today’s article! As always follow me on here or on Twitter @thant1194 in order to stay up to date on all my articles as I release them. I am not associated with the Fetch.AI team and at the time of this publishing am currently a holder of their token. Thank you again for reading. I have to return some video tapes.