So what is the traditional argument for how artificial intelligence can provide value in areas that have been traditionally occupied by human intelligence, for example, general decision-making? Well, the first thing is lower cost and the ability to operate at scale and operate efficiently. For example, well-known human biases that include order effects. In other words, humans make different decisions for the same challenges or questions based on the order which they see them, effects related to time of the day or mood. The impact of individual experiences. For example, a doctor gets into a crash on the way to the office. It turns out that can affect decision-making. Well, algorithms ostensibly not only allow scale and speed, but it regularity that would otherwise not necessarily be in place for human decision-making. Machines may not be only less error prone than humans, but could do all of the above faster, and again at greater scale. Another fascinating area where artificial intelligence has seen success and use cases is automated customer support. Some of you might have had the experience in which either through a chatbot online or even speaking on the phone, you've encountered algorithmic counterparts or decision trees that appear to take a human character when you interact with them. Having a personal touch or the ability to be relatable. It's clear to see how this might lower cost because the marginal cost of an algorithm might be nearly zero, whereas the marginal cost of a person maybe related to wages, and benefits and so on. In addition 24-7 availability may alter the case for algorithms in customer support for the rest of time. Now, there are some controversies. Clearly, we see in the robo advisory industry applications involving AI and a future for AI. However, it's not so clear exactly what the role will be in equilibrium, and in fact, if whether the equilibrium is stable or constant at all. There have been some who have suggested that the entire investment advisor industry will move away from humanity and toward algorithms, interaction with chatbots or robo advisors, and ultimately disintermediate humanity. In part due to legal and compliance issues however, at least currently, it's unlikely in the near term that acute computer can be or will be held to a fiduciary standard which is required in most countries including the US by laws. For example, the Investment Advisors Act of 1940 requires a fiduciary duty or the Employee Retirement Income Security Act, ERISA which governs pension related activities and some cases requires a fiduciary duty, which is not clear algorithms can arise to. This controversial area will certainly be addressed in the next several years.