Artificial intelligence technology is gradually integrating into many areas, including finance and loans. Yoram Kraus, a high-tech entrepreneur who focuses on human-machine interaction, explains how artificial intelligence could improve processes, bringing about a true revolution in loans and financial services.
Yoram Kraus, a high-tech entrepreneur working in the field of artificial intelligence, has followed a very different path from that of other successful high-tech entrepreneurs. Having completed his military service as a field reconnaissance officer he founded a real estate development company which built dozens of projects in Israel and around the world. Over the years, Kraus has developed an expertise in real-estate projects for the high-tech industry, building offices for tech giants such as Google, SAP, and Melanox, to name a few. The high-tech industry has always piqued his interest; about a decade ago, Kraus began his activities in the start-up industry, and even launched a number of high-tech companies in the fields of algorithmics and smart systems. In recent years, Kraus’ activity has been focused on artificial intelligence and human-machine interaction, including possible AI applications in finance and investments.
“One of the tools I have been looking into for some time now is an artificial intelligence system in the field of human-machine interaction which makes it possible to extract emotions and to characterize the digital-world user by analyzing mobile phone usage characteristics. Using smart tools and advanced systems we want to crack the code of human motivation. Once you know what drives people you can use this knowledge across many different verticals. One such area is finance.
“Today’s systems have reached a glass ceiling, because they collect data from the Internet, upload them to the cloud, carry out statistical analysis on the big data which they gather, and derive behavioral patterns. This is actually where the work of AI ends: it produces superficial analyses and characterizations compared to the potential that I investigate.”
“In finance, and especially in lending, there are two main parameters which the lender must look into. One is the default rate: what is the proportion of people who may not be able to repay their loan. The other is the loan denial rate, which is the proportion of loan applicants whose loans wouldn’t be approved from the start because an analysis of their financial situation has shown that they would be unable to repay their loan and they therefore do not meet the criteria for getting the loan. The loan denial rate usually ranges between 80%-90% of loan applicants. The lender would rather give fewer loans, but to do so knowing that those clients who were given a loan would repay it with certainty.”
“The technology currently under development is going to create an additional layer of loan applicant characteristics – the psychological layer. This is a new dimension that would provide a clearer and more accurate picture of the borrower. This picture would answer questions such as: is the user about to commit fraud? Do they know how to manage their finances in a way that would allow them to repay their loan? These data cannot be extracted through statistical behavior – the foundation of today’s artificial intelligence. We would do this using symbol artificial intelligence (SAI) tools, and then apply the model by analyzing mobile phone usage behavior.”
“SAI technology will be the focal point for developing the next layer of data systems worldwide. This signifies a transition from statistics-based systems to systems that can analyze individual human behavior. The new system produces data on each individual user, and generates a unique psychological profile of the user. This is something no system is capable of doing today, since present-day artificial intelligence deals with statistics, with big data analysis, and with portraying social behavior patterns.”
“The advantages proposed by an SAI system stand out particularly with regard to credit fraud prevention. There are three factors at the base of every financial fraud: the borrower’s personality traits, their motivation to act, and loose controls and lacking regulation. Currently, no system is able to analyze the first factor. Systems check borrowers’ motivation only on a superficial level based on big data and not individually – this is a binary, rather than holistic, analysis. SAI systems can lay out these two factors which are most critical for protection against fraud and may, further down the road, lead to a reduced rejection rate. This is very significant because it may allow lenders to increase profits at tremendous rates. Lenders pay millions of dollars per year for the data they buy. They pay dozens of dollars for each user, and in order to obtain statistical insight they must analyze a lot of data and history. These expenses add up to millions of dollars for information which, at the end of the day, is only statistical information.”
“SAI systems analyze individual user-level data independently and automatically, allowing credit companies to pay a few pennies for each individual user who, to begin with, is likely to receive a loan. This allows us to increase the conversion rate. The significant changes expected to arise from using SAI systems are wide-ranging, both for individual users who may now be able to get a loan, for the bank which would make greater profits, and for the government which benefits from more people being able to carry out transactions. Our impact on the economy is huge, and that’s what drives us to do what we do.”
“Financial services is just one of the numerous fields which artificial intelligence is going to improve, if not drive forward. When I travel around the world and meet high-tech entrepreneurs and executives, they often ask me how I ended up in artificial intelligence. I tell them that other than being an entrepreneur, I have been mountain-climbing for nearly 20 years, with dozens of climbing expeditions in Europe and in Asia under my belt. One time, when I took my youngest daughter with me on one of the climbing expeditions, as we reached the peak she asked me what would happen if I were to climb some mountain and not come back. I said that I would make her an avatar in my image, which would look like me, talk like me, sound like me, and most of all – it would think and feel just like me. That led me to artificial intelligence. We are on the cusp of a fascinating time. A true revolution where we learn to teach a machine how to feel. We are still at the beginning, but the possible applications are endless, and they will change the world and improve the lives of millions. As I said, a truly exciting time.”