How Exactly AI and Data Science Can Replace a Human in the Financial Sphere How Exactly AI and Data Science Can Replace a Human in the Financial Sphere
When talking about Artificial Intelligence, most people immediately imagine armies of robots, which will definitely revolt against humanity. Yes, we are... How Exactly AI and Data Science Can Replace a Human in the Financial Sphere
When talking about Artificial Intelligence, most people immediately imagine armies of robots, which will definitely revolt against humanity. Yes, we are all influenced by science fiction and movies, but we should separate myths from reality. Step-by-step AI is changing the world. Is it bad or good? Should we worry about our jobs? What can we expect in the nearest future? Let’s inspect the financial sphere as an example.

What can be classified as Artificial Intelligence in finance?

[Related Article: A conversation with Thomas Wiecki on the use of probabilistic programming and machine learning in quant finance.]

Originally, the idea of Artificial Intelligence was based on an assumption that all our human thoughts, understanding and analyzing, perceptions and conclusions could be mechanized. But the theory, that all rational thoughts could be systematized by algebra or geometry, was refuted for almost century ago. Scientists showed us that mathematical logic has serious limits and cannot cover all kinds of human thoughts. But there was a bright side of this statement – within these limits any form of mathematical thinking could be mechanized. Today’s Artificial Intelligence is based on the ability to operate large amounts of data. Basically, it analyzes, answers questions, and solves problems using certain rules in the specific area. But unlike humans, AI can operate a massive of information or Big Data, how we call it. Neil Jacobstein, the head of the AI and Robotics Track at Singularity University on the Nasa Research Park said, Artificial Intelligence and Data Science in finance are “utilizing data, specialized hardware, and machine learning algorithms to augment humans and allow people to utilize this technology to do things they never imagined that they could do before”.

So, AI helps us solve business and technical problems. It provides an opportunity to delegate tasks to computers, which previously only a human could do. Why is this important? Technologies are able to avoid huge mistakes, reduce costs, improve efficiency and security, which is especially relevant for finance sphere. For example, large companies with $500 million of annual payments could lose up to $2.5 million on mistakes and cheating, asserts Forbes. But AI can expand monitoring and analysis and prevent this cash leakage.

How market responds to the changes

Globally, the market is preparing for these changes, but slowly and carefully. According to the Global FinTech Report 2017, 30% of large Financial Institutions are going to invest in the development of AI technologies in the next three to five years, but 82% of them are ready to expand partnership in this sphere. This market growing tendency was also confirmed by McKinsey Global Institute report. It says the most attention on new technologies will pay businesses, governed by predictions, fast decisions, and personalized customer connections. There are clear benefits from the implementation of AI systems in financial services, and the report predicts the market growing up to $3 billion by 2020.

In which directions is FinTech able to grow? First of all, mobile banking. Smartphones are entering more and more aspects of our lives, and mobile banking and shopping with the support of chatbots will reliably root in the market. Business Insider forecasts mobile commerce will reach 45% of global e-commerce sales in the next two years. Mobile banking has huge potential in the third world countries. In some of them, India or Venezuela, for example, the level of currency devaluation is so high, people need to go to the market with stacks of cash. In consequence, investors pay more attention to cryptocurrencies, free from government control. Also, in these countries penetration of local banking is more than 10 times lower than mobile devices, which gives the green light to mobile banking.

[Related Article: Watch: Introduction to Quant Finance with Quantiacs Toolbox]

During the third quarter of 2017 financial technologies, in general, gathered over $8.2 billion investments worldwide. The biggest improvement is expected in cyber-security and payment intelligence, but especially in customer experience and facial recognition technology, where AI is a critical tool. Furthermore, Gartner predicted 20% of adult people in developed countries will use AI for everyday tasks by 2020.

How people and robots will share the area

The basis of the finance industry is processing information, of course, so it is not a surprise many key operations are already digitized. New software takes into account all necessary information about the client’s credit background and even more, so robots can make independent and reliable decisions about loans. Stockbrokers and financial advisers could be replaced by robo-advisers by making personalized investment portfolios. And, according to The New York Timesnumbers of financial analysts are already replaced by software. For example, one bot can replace three to five workers and save a lot of money for the company. But the great majority of finance structures prefer to digitize only the most routine office jobs and let the staff do more interesting tasks. A professor at the London School of Economics Leslie Willcocks said: “It takes the robot out of the human”.

In 2013 Oxford academics released a report on the future of employment. They predicted 47% of current professions will be automated during the next 20 years, and 54% of financial jobs were claimed as a “high risk”. But even one of the paper’s authors, Carl Benedikt Frey, thinks that if half of the jobs disappear it does not mean, that half of the people will not have jobs. New technologies kill old professions as well as create new ones. There is a probability that new jobs will appear much slower and at the same time, we will have a lack of professionals in new narrow specializations. Here we will need support from universities. In the UK, Wrexham Glyndwr University has already launched the first in the country undergraduate degree in FinTech. In the US, business schools in both Stanford University and Georgetown University are going to offer new FinTech direction in their MBA programs. There are still drawbacks in these proses, like the lack of academic textbooks or professors, but it is the first step, and there will be more to come.

What changes can already be seen

Neil Jacobstein mentioned nine main tasks for new technologies in the financial sphere:

  • Market Research
  • Accounting
  • Acceptance
  • Investments
  • Credit
  • Insurance
  • Collections
  • Predictive Analytics
  • Compliance

Market Research gives financial companies an opportunity to ask the questions they have never asked before and get the answers. Big companies like Data Miner are not orientated only on FinTech but allow financial structures to get all necessary information. In Accounting, there are companies like Zietgold, which provides AI-based accounting systems on clouds or even on smartphones. These systems allow all financial operations to run on a smartphone for businesses as well as independent contractors and personal transactions. The great example of Acceptance is Amazon’s Alexa. It is not specialized on finances, but its skills grew exponentially from 135 in 2016 to over 10,000 in 2017, and many of them financial. Or Digit – the program, which helps people save their money by providing a personal forecast of the cash flow.

As for Investments, there is a big company Sentient, a part of the team which built Siri, a virtual assistant for Apple. The team made a groundbreaking platform, which creates small agents. They behave as hedge fund managers, can make predictions and even evolve in competing with each other. Credit is another big area in FinTech. One of the most important directions is to make loan process more transparent and clear, especially for the younger generation. Affirm is a good example of the companies, who already realized that ideas. In Insurance AI is using to collect information and validate the claims that people make for insurance, as, for example, Lemonade does. The company is using AI in all aspects of work from sales to making insurance decisions. In Collecting area finance structures like CollectAI makes AI-based collection agents, which can immediately learn what works in the different situations and do it much more productive than human. Or, for example, Accord company, based on the idea, that sometimes collection problems are result of misunderstandings or lack of knowledge. Their program is analyzing the situation from both sides for they were able to come to a common accord. Predictive Analytics goes to a new level. Opera Solutions, for example, provides data analysis services based on signals – mathematical transformations of data. It creates a repository for the synthesized intelligence, which is a mediator between data and people who use this data for the better understanding the situation and making the best decisions. As an example of Compliance, we have a Trifacta, a result of the partnership between universities Stanford and UC Berkeley. The software is cleaning the data, getting out the noise to improve the productivity of people, who are working with data.

As a conclusion, Neil Jacobstein convinces, with AI solutions in financial sphere we can get best-in-class performance for some tasks, improvements in speed and prediction. These technologies do not need vacations or sick leaves but have serious drawbacks – limited empathy, language understanding, and social grace. So, AI technologies are really able to make our lives more convenient and easy, but there are always will be need of people to control, improve, and guide the machines where we need to go.

Originally posted on Ralabs.org



We solve challenges for Fintech industry using data science and machine learning. We are Ralabs, your reliable partner in the ever-changing world of Financial Technology. https://ralabs.org