Robots have arrived on Wall Street, and humans need to learn how to work with them. JPMorgan Chase & Co. is rolling out virtual assistants to maintain help desks, track errors and route inquiries. A Bank of America corporate client made a $1.6 billion payment by mobile phone. AllianceBernstein Holding LP’s robot, Abbie, can gather 4 million data points a day and execute bond trades directly with counterparty bots. U.S. banks could axe an estimated 200,000 jobs over the next decade with bank branch, call center and corporate employees bearing the brunt of those cuts as lenders automate back-office functions, according to Wells Fargo & Co. But all hope is not lost.

1. What makes robots qualified for Wall Street work?

Artificial intelligence, a branch of computer science that aims to imbue machines with aspects of reasoning. Today, the term includes machine learning, which is the ability for computers to learn by ingesting data, and natural language processing, the ability to read or produce text. What’s known as robotic process automation is a simpler form of AI that performs rote tasks like answering administrative requests.

2. Who is most at risk of losing their job to a robot?

The first to go will be those with roles that are repetitive in nature: support functions, back-office processing, producing reports that rely on structured data. More than 50% of tasks performed by loan officers, financial advisers, bank supervisors, loan clerks and tellers could be automated or augmented by technology by 2025, according to Accenture. In capital markets, more than half of the work done by financial analysts, sales agents, brokerage clerks and statisticians could be automated or augmented. Headcount for more prestigious front-office workers -- such as traders and investment bankers who produce revenue -- could drop by almost a third, according to McKinsey & Co. JPMorgan Chief Executive Officer Jamie Dimon said the company is evaluating which of its jobs are most susceptible to being lost through AI so that it can retrain its workforce.

3. Who else should be worried?

Consider the junior investment banker, who spends much of his or her time collecting and analyzing data and then creating reports. Even the highest-paid jobs for portfolio managers could disappear as firms use data scientists to solve some of the most complex investment problems, according to Marcos Lopez de Prado, a Cornell University professor and the former head of machine learning at AQR Capital Management LLC. Many of the 6.1 million people employed in the finance and insurance industries will lose their jobs because they aren’t trained to work alongside algorithms, Lopez de Prado said. In a survey of CFA Institute members and candidates, 43% said they expect their roles to change significantly in the next five to 10 years. The three roles most likely to disappear are sales agents, traders and performance analysts, they said. For the buy-side traders who already use automated trading, more than 20% of corporate bond deals now happen with little or no human interaction, according to Greenwich Associates.

4. Is there good news for anybody?

New jobs like machine-learning engineers and data scientists are increasingly in demand. Wall Street’s technologists are gaining prominence and their departments command ever bigger budgets. Finance’s junior ranks are swelling with recruits trained in quantitative disciplines such as mathematics or science.

5. What’s the strongest case for keeping humans at the wheel?

It’s still hard to automate empathy or trust, so as long as clients want to phone a salesperson, people will have jobs. Transactions that are large, complex or take place in illiquid markets are harder to automate. Computers are learning to understand bond covenants and court documents, but contract interpretation remains a challenge. Still, the humans who remain in the industry will need to “unlearn” their expectations about bonus culture as machines take over.

6. How should those starting out in finance prepare?

Take a hint from what the banks are saying: Citigroup Inc. plans to hire 2,500 programmers this year. It’s also teaching its traders to code, while making sure its coders know how to trade. Goldman Sachs Group Inc.’s trading division is also adding engineers. Learn to work with the tech as it’s deployed across the business. Focus on relationship management, which probably can’t be done by computers. Or focus on data science, helping banks analyze the information that AI needs. Learn Python.

To contact the reporters on this story: Lananh Nguyen in New York at lnguyen35@bloomberg.net;Jenny Surane in New York at jsurane4@bloomberg.net

To contact the editors responsible for this story: Michael J. Moore at mmoore55@bloomberg.net, John O’Neil, Daniel Taub

©2020 Bloomberg L.P.