The idea that robots have the potential to eliminate millions of blue-collar jobs is nothing new. In February, for example, Changying Precision Technology Co., a mobile-phone manufacturer in Dongguan, China, laid off 90 percent of its staff and automated its production lines. Only 60 people now run the factory, which used to employ about 650.
But robots also are coming after high-paying positions that have traditionally required years of education and training. Here are a few examples.
In 2000 Goldman Sachs, a multinational finance company with headquarters in New York City, employed 600 equity traders. Today there are just two of these traders left. Automated trading programs do the rest of the work, supported by 200 computer engineers, according to an article in MIT Tech Review. Marty Chavez, the company’s chief financial officer, said that stock trading and investment banking will be the next thing to become computerized and automated.
Stockbrokers make on average about US $72,000 per year, although brokers working on Wall Street tend to make more. But average compensation for workers in equities sales and currency trading—which also can be automated, according to Chavez—is closer to $500,000. Those figures are from Coalition, a research firm in London that tracks the financial industry.
Babak Hodjat, an IEEE affiliate member and cofounder of Sentient Technologies, an artificial-intelligence company in San Francisco, told Bloomberg that humans are too emotional for the stock market. “It’s scarier to be relying on those human-based intuitions and justifications than relying on purely what the data and statistics are telling you,” Hodjat said.
His company is developing an AI program that can scour data, spot trends, and learn to trade stocks. At this point, the company has completely automated investments for its own hedge fund. There’s no word yet on how well the fund is doing, but the AI program could pave the way for other hedge funds to follow suit. And that’s a bad omen for hedge fund managers, whose starting salaries can be as high as $100,000 per year.
For better or worse, patients and hospital administrators might be dealing with fewer humans when filing an insurance claim. Fukoku Mutual Life Insurance, in Tokyo, replaced 34 of its workers in January with an AI system designed by IBM.
The system, IBM Watson Explorer, scans hospital records and other documents to determine insurance payouts based on types of injury or illness, patient medical histories, procedures administered by doctors, and length of hospital stays, according to the company’s news release. Automating those research and data-gathering tasks helps the remaining employees process the final payout faster, the release says.
The system costs about $1.7 million plus nearly $128,000 per year to maintain, according to an article in The Mainichi. Fukoku, however, anticipates saving roughly $1 million annually on employee salaries. Additionally, Watson is expected to improve the company’s productivity by 30 percent.
Lawyers and their assistants, who often spend long hours poring over thousands of documents to prepare for cases, are likely to feel the impact of automation, too. AI platforms are being developed to mine documents that are used as evidence in trials, create and review contracts, and identify potential fraud.
One platform, LawGeex, helps lawyers create and review contracts. LawGeex takes a new contract, scans it, and compares it to a database of similar contracts. The program eventually could learn to create new contracts and flag errors in existing ones.
The LawGeex company founder and CEO, Noory Bechor, is a former lawyer who says he wants to save lawyers time and money. “For me, it was mind-blowing that I needed to reinvent the wheel each time I needed to create a contract or each time I needed to review a contract,” Bechor said in a CNBC article.
Another new AI platform is eDiscovery, which searches documents for information relevant to lawsuits and other litigation. It was developed by NexLP, an AI software company in Chicago. The company’s founder and CEO, Jay Leib, acknowledges in the CNBC article that the software can miss things, but he says it’s not as fallible as a human. He adds that the sheer amount of data generated today makes it hard for workers to keep up.
The success of such AI tools probably means that paralegals—who earn up to $100,000 per year—will have fewer job opportunities in the future.