Artificial intelligence is a term that is frequently used when discussing the most cutting-edge developments in technology today. Artificial Intelligence or “AI” is a broad term that is defined as “the theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages” (Oxford). This means that AI is applicable to diverse fields including, computer audio or visual recognition, self-driving vehicles, autonomous robots, film recommendations on streaming, or financial analysis. What appears to signify an ‘artificially intelligent’ system is some degree of autonomy, the ability to perform a task without constant human intervention, and a level of adaptability, meaning a system that has the ability to change outputs, bettering its own performance, by learning from its own experiences.
The capability to adapt mentioned above illustrates another very common term , “Machine Learning.” Machine learning enables systems to use algorithms in order to increasingly better their performance. The algorithms make predictions on inputs and are able to adapt their predictions based on the results. A simple example would be the use of spam filtering tools built within an email system. The algorithms observe occurrences of spam messages, and are able to make predictions as to whether incoming email messages constitute spam or not, based on what it has previously observed and marked as spam.
AI can be categorized as either weak or strong. Weak AI, also known as narrow AI, is an AI system that is created and trained for a particular task. Virtual personal assistants, such as the iPhone’s Siri, are a form of weak AI. Strong AI, also known as artificial general intelligence, is an AI system with generalized human cognitive abilities. When presented with an unfamiliar task, a strong AI system is able to discover a solution without the need for human intervention.
Deep learning is a type of machine learning that models patterns in data as complex, multi-layered networks.(Martin Heller). Deep learning is the most general way to model a problem and it has the potential to solve highly difficult problems—such as computer vision and natural language processing—that outstrip both conventional programming and other machine learning techniques.
Investing in AI is an international trend that brings a satisfactory return. The number of investment transactions grew globally, from less than 200 investment deals in 2011, to over 1,400 in 2017(OECD). This corresponds to a 35% CARG from 2011 into the first half of 2018 (See below). Startups located in the United States attracted a crutial portion of all investment deals, rising from 130 in 2011 to almost 800 in 2017. The EU has also seen a significant increase in the number of deals, from approximately 30 in 2011 to around 350 in 2017.
According to Nasdaq Global Information Services, the EurekaHedge indexes’ performance shows that, for funds covered by the index, the AI/Machine learning hedge funds outperformed the average global hedge fund for all years except for 2012. Besides, while returns have been more volatile compared to the hedge funds covered, the AI/machine learning funds have posted considerably lower annualized volatilities compared with systematic trend following strategies.
It was also reported that the EurekaHedge AI/Machine Learning Index posted better risk-adjusted returns over the last two- and three-year annualized periods compared to its peer indexes described here, with Sharpe ratios of 1.51 and 1.53 over both periods respectively.
Authors: Derek Frappa & Enrique (Ling) Wang, Pycap Interns