Artificial Intelligence in the Crypto Industry — How Much Power Yet?

The key to artificial intelligence has always been the representation.

Jeff Hawkins

The real implications and use cases of artificial intelligence (AI) are yet far less than they are thrown around in the fintech world. When we talk about AI, people generally tend to associate it with almighty “cyberpersonalities” who know literally everything and have outstanding power, or with an utopian world where machines rule humanity. Well, suffice to say, that’s a bit of an exaggeration of the technology.

Basically, AI is a methodology through which machines can be trained to mimic and execute specific tasks that have the best chance of achieving a specific goal. This is considering that the AI has access to enough historical data to learn from.

For example, we’ve been watching the growth of AI in self-driving cars. In this case, the machine is trained to acknowledge all the external data through a range of sensors and cameras and process it in a way to avoid a collision. Thereby, it measures the consequences of any action it takes, as each action will impact the end result — to avoid a collision.

Now that we’ve settled on the fact that AI is not going to become a machine overlord for humans (at least in the foreseeable future), let’s get into the relevance of AI in cryptocurrency. Similar to AI, the blockchain technology and cryptocurrency is poised to revolutionize finance as well as tech space. Currently, there are more than 90 cryptocurrency projects that are connected to the AI industry and holds a market capitalization of over $740 million in total. With so much money flowing into this ecosystem, let’s take a look at the actual use cases of artificial intelligence in blockchain technology.

Interaction of Nodes on a Blockchain Network

In a blockchain network such as Bitcoin or Litecoin, all the nodes continuously interact with each other and use the Proof of Work (PoW) consensus model to validate new blocks. However, this consensus model has major drawbacks such as high energy consumption and vulnerability to 51% attacks. Recently, we saw two such attacks on the Ethereum Classic network — both in a span of one week. Although no system or technology can claim to be 100% secure, with the help of AI, a system can be trained to predict authenticity of transactions by evaluating the past behavior of the system.

For example, an ambitious project called Fetch.AI has come up with a unique consensus algorithm called Useful Proof of Work (uPoW). It offers several benefits over the traditional Proof of Work algorithms using the AI. Unlike in the traditional PoW model where nodes must download every block and add them to the chain sequentially, Fetch.AI with its uPoW will deem any transaction valid once it’s confirmed by two nodes leaving computational resources free to train the AI. Feth.AI also uses natural language processing and machine learning algorithms to train the network from being manipulated by scammers to run fraudulent transactions.

Processing Big Data

Similar to Fetch.AI, another promising project called DxChain also seeks to employ machine learning and big data to store, trade, and compute big data within a decentralized framework. For the consensus model, it uses the Proof of Spacetime (PoSt) protocol and aims to design a platform to solve the problems of computation of big data in a decentralized manner with the help of AI.

In 2017, data surpassed oil as being the most sought-after commodity, and it’s no surprise that the importance of controlling your personal data has found its way into the blockchain tech. Currently, only big companies have the capability to run these big data tasks because they require expensive hardware. Another reason that these companies own most of the existing consumers’ data.

DxChain claims to change the big data processing using the distributed ledger technology powered with artificial intelligence. Currently, the well-known blockchain networks such as Bitcoin or Ethereum are turning out to be a bottleneck as they lack the computational ability to process big data or provide storage services. Unlike projects such as Lightning Network, DxChain doesn’t believe in building a second layer on Bitcoin and only increasing the Transactions per Second (TPS). It rather takes an approach of making blockchain network function as a computing unit. Basically, DxChain aims to workout both the storage limitations & computational ability of the current blockchain networks simultaneously and has a vision of embedding blockchain technology in every application area of the internet.

Automated Trading

Automated trading is probably the most mainstream use case of artificial intelligence in cryptocurrency. Within the last few years, especially in the past few months with COVID-19 pandemic, cryptocurrency trading is growing a popular appeal. However, the unpredictability and volatility of cryptocurrencies continue to put crypto traders’ investments at high risk. Also, when it comes to financial markets, it’s practically impossible to prevent human emotions from affecting important trading decisions.

And here come AI-based trading bots. These bots can be trained to act solely based on historical data to make informed trading decisions without being overruled by human emotions. There are multiple pieces of research proving AI can help the cryptocurrency price prediction model with precise accuracy. However, the efficacy of AI in each case depends upon the nature of the involved machine learning model and the sample set sufficiency. Major exchanges such as Binance and Kraken already let traders use their API allowing them to use third-party trading bots.

Brave New World

These were several use cases of artificial intelligence in reference to blockchain technology and cryptocurrency. Apart from the projects that we’ve covered, there are others — like

  • Ocean Protocol — helps developers build Web3 (decentralized internet) apps.
  • Velas — a self-learning blockchain ecosystem claiming to improve scalability and security using AI-powered Delegated Proof-of-Stake (AIDPOS) consensus mechanism.
  • Numeraire — a crowd-sourced hedge fund powered by AI.
  • Project Pai — another ambitious project claiming to create an intelligent 3D avatar for everyone — Personal Artificial Intelligence (PAI),
  • and many more.

Some of these projects are quite ambitious and it would be interesting to see their real-life implications on future generations. In the near future, we are expecting to see more projects superimposing the two greatest tech ideas of the 21st century — AI & Blockchain.