What is Fetch.ai?

Last week, it was announced that Binance had completed its’ second successful major ICO via its’ new Launchpad service. The chief reason that bears mentioning here is simple. The ICO in question was being run by a project called Fetch.ai, which seems to be truly unique in its’ combination of different technologies to achieve a highly ambitious goal.

First, Fetch.ai has combined the technical framework of Directed Acyclic Graphs with both Proof-of-Work and Proof-of-Work consensus protocols. If you’re scratching your heard at this point wondering how all of this works out, then you’re surely not alone.

First and foremost, it appears that this means the team has created a digital ledger that uses all of these technologies at once. Clarifying this any further will require time-tested usage of the platform, though the team has stated that a DAG and not the Blockchain, is where the network’s transactions are stored.

Next, Fetch.ai has the unique vision of creating a machine-driven economy, which is where it is truly easy to get lost in the weeds. Do they mean that they want to induce some sort of trading of goods on a wide scale that does not really involve humans at all? Do they mean something else entirely?

The easiest way to answer these questions is to say that the future Fetch.ai ecosystem aims to be a bridge forward. In an interview with the WhatBitcoinDid podcast, Fetch.ai’s CTO, Toby Simpson, clarified why this is the case. Overall, according to Simpson, the Fetch network aims to be the foundation for both the Internet of Things and the Economy of Things, simultaneously. Another way of conceptualizing this is to say that the Fetch network will be where machines share data with each other and transact with each other, non-stop.

The central technology that makes any of this is possible is what Simpson and the Fetch team call “autonomous agents.” Reportedly, these are entities that can transact with each other in just about any context on the behalf of humans. To understand why this is important, it’s essential to consider a specific case of how this might play out.

Imagine that you own one of Tesla’s Powerpacks or Powerwalls. What if your electric vehicle charging stations was connected to a network that knew how much energy you needed for your car and how much was usually left over, in a certain period of time. In this case, apparently, with Fetch, you would be able to sell your excess solar energy without spending any time out of your day to do so. How this process plays out is simple. According to Simpson, initially, you would input certain data points that allow the Fetch network to draw up an autonomous agent, which you then allow to be constantly tied to your charging station. At this time, it seems as if there would be a considerable amount of help from the team throughout the process since they aim for their agents to be something that anyone can bring online and use in any situation.

With that in mind, it is definitely reasonable to wonder: is this the dawn of AI systems doing all of our daily business for us? Will the Fetch network enable all of my purchasing of any kind to be done without me?

On the one hand, Simpson made it clear that this is the goal. Fetch wants to bring about the age of a fully-functional and reliable Internet of Things and at the same time, bring about an age where machine learning algorithms take away most of the stress from our daily lives. Think about how annoying it is to find the right travel package. What if you could give all of the relevant information in that regard to an agent that would do the work for you? There would be almost no more stress on paper, as the agent(or algorithms) would be acting on your behalf with your needs in mind.

Even though this is a possibility, issues still remain with this idea. For example, how does the Fetch team plan to get buy-in from businesses for autonomous agents to act as travel agents? Even more importantly, why would businesses want these agents to act on their behalf during the same process? Perhaps, it all boils down to trust.

If Fetch can clearly communicate why trusting in the code of their machine learning algorithms is better than trusting in humans who have natural human biases, then everything just might work out for them. Simultaneously, if anything falls apart in this regard, their solution just might fall flat. What already works against them is humanity’s natural distrust for AI, which is often fueled by the fear-mongering nature of the global media. Making matters worse for Fetch is the added point that it may seem as if they are trying to take away the human element from purchasing altogether.

Therefore, while their recent ICO on the Binance Launchpad seems to have gone well, they will have to appeal to the greater public beyond the crypto-sphere if they ever want to see their age of autonomous agents come about. What may help them here is the fact that their agents are not truly autonomous. As mentioned above, they are basically Machine Learning algorithms that take in an initial set of training data from humans and input it until they reach the goal that it specifies. Considering this, Fetch’s central message could be the easing of people’s difficulties instead of the taking away of their power.

If you’re interested in learning more about this project, take a look at some of the sources below that we have compiled for you.