Original article was published on Artificial Intelligence on Medium
AI and ML, call them mere conundrums or established axioms, there will be multiple proponents from varied fields that would readily vouch for the value they continue to bring in. Most of us don’t realize it but these technologies are a greater part of our lives now than we can possibly fathom. Virtual personal assistants, navigational predictions, video surveilling, face recognition in social media services, email spam and malware filtering, customer service chat bots, refined search engine results, are just a few of a plethora of touchpoints. The curiosity and buzz around these inventions are still in their nascent stages and technophiles are laying their focus on identifying opportunity areas and decoding intriguing potential future applications. It was in this setting that experts and companies recently took to embracing the convergence of AI, ML and blockchain. This tripartite coalition, due to its resulting synergy is now gaining traction exponentially.
TraneAI is leveraging blockchain’s property of distributed computing to speed up the machine-learning process and consequently eliminating the need for a centralized infrastructure. Then there are softwares like DeepBrain Chain and SingularityNET, again based on blockchain’s principle of decentralization, develop cost-effective and easily accessible neural networks to power AI efforts. With the cryptocurrency markets being highly speculative and volatile on one hand and the world witnessing an unprecedented growth in the adoption of blockchain and distributed ledger technologies (DLT) on the other, there is huge potential of AI and ML to be leveraged in these markets. Santiment, is one such tool that collects and sells live markets data feeds, by employing AI and ML, to crypto traders. On a very high level, investors in the crypto markets are benefitting hugely from tools like NeuroBot, Senno and ThinkCoin, that aid in quick and timely technical analysis of large volumes of market data with high efficiency. Dig deeper and one can discover more intricate and fundamental ways in which technologies like AI and ML can solve the crypto market enigma.
Persona matching is an interesting application of blockchain for cryptocurrency trading, crypto-to-fiat and fiat-to-crypto transactions. Cryptocurrency being a highly volatile asset, is often characterized by unstable liquidity. Due to this, there are only a handful of merchants that accept crypto as a mode of payment. Such merchants, liquidity lenders or even banks, more often than not, levy a heavy surcharge to compensate for the high volatility and liquidity gaps. Coinsbank, a crypto exchange, for instance is a major provider of crypto-to-fiat transactions and starts from $50 as transaction fee for every transfer. To deal with this, TradeConnect, a ThinkCoin-powered trading platform employed blockchain to enable trade and exchange between traditional and crypto assets. This platform, based on the tenet of persona matching, identifies the optimal combination of financial assets and traders. To explain it further, the platform generates a digital persona for users and tries to find a broker or bank that would be its best fit based on the user’s trading activities and the lender/broker’s price quotations, in turn benefitting both parties.
Furthermore, AI has already proven to be more capable at decision-making than human beings. While it is almost inevitable for humans to decouple decision making from their emotions and biases, machines as we know them, never encounter these constraints. AI-based automated trading is being carried out by technologies that are pre-programmed with entry, invest, divest and exit conditions for taking trading decisions. With this, decision making is devoid of even the slightest traces of irrationality. According to Bitcoin.com, Renaissance Technologies’ Medallion Fund is a relevant case in point. It was reliant on AI for taking trading calls and during the 2007–08 global financial crisis, this fund gave returns of over 85% annually.
Trading in financial asset markets can be made profitable by exploiting the market inefficiencies and information asymmetry. Sentiment analysis through data crunching, hence, becomes a key game-changer. AI and ML are already being leveraged to analyze large tracts of textual data from blogs, articles, social media posts, etc. to ascertain a market’s sentiment on any political, global or economic development. Senno is one such platform. It is also blockchain-enabled in order to make the market sentiment-analyzing algorithm accessible to a larger audience. In relation to the crypto market, Senno has come together with CryptoScanner to scan various cryptocurrency channels and create reports and alerts, showing the market’s views of individual coins. Traders then use these insights to chart out their strategies, now being better-informed.
On similar lines, algorithms enabled by AI and ML are also being used to make more accurate forecasts and predictions. Augur, is a platform that combines such algorithms with blockchain to make it more accessible. NeuroBot is another tool that makes use of neural network for speedier and more precise price predictions. The algorithm is used to crunch data perpetually and arrive at probable future prices that are claimed to be accurate 70–90% of the times. It takes into consideration essential technical analysis like patterns and signal indicators like the Fibonacci retracement and Elliot wave that also capture trends in investor psychology, among other factors.
Apart from those already cited, there is a slew of platforms that are capitalizing on the amalgamation of AI, ML, blockchain with cryptocurrency trading and this is undoubtedly becoming a promising industry in itself. AdHive, Cryptoindex, MATRIX, Sentigraph and Peculium have gone on to become few of the most successful ones in the last couple of years.
Hence, it can be established that AI, ML and blockchain are proving to be irresistible to cryptocurrency investors. However, it is also pertinent to be cognizant of the flipside. It is believed that AI-fueled version of wash trading caused the illusion of increased market activity and believed to have driven up the price of bitcoin on the Mt. Gox exchange between November and December 2013. Rogue trading algorithms called trading bots named Willy and Markus bought 650,000 bitcoins on Mt. Gox that eventually led to a bubble-like situation creating artificial scarcity, consequently leading to its collapse and subsequent bankruptcy. Regulation can bring about some vigilance but it brings with it its own set of cons.
All in all, while there is no denying that these technologies have proved to be disruptive in terms of providing deep insights distilled from vast amounts of information, it is still expected of investors to temper their expectations and exhibit guarded faith in them. Only time will tell if these technologies will really be able to crack the crypto market fully and also guarantee responsible and rightful implementation.
Aitken, R. (2018, May 31). Can The ‘AI Blockchain’ Combo Finally Crack The Crypto Market? Forbes.
Haverans, R. (n.d.). BrainBridge. Retrieved from BrainBridge: https://www.brainbridge.be/news/how-ai-and-ml-are-changing-the-game-for-cryptocurrency-investing
NewsBTC. (n.d.). Retrieved from NewsBTC: https://www.newsbtc.com/2018/06/25/bitmain-raises-400-million-prior-12-billion-ico-growing-dominance/