When you read or hear news stories about the imminent takeover of robots and algorithms that will eliminate jobs for human workers, many times the first examples given are blue-collar jobs like factory workers and taxi drivers. And you may have mentally congratulated yourself because your “professional” job is safe from the threat of being outsourced to computers.
But don’t feel so safe just yet. More and more, sophisticated algorithms and machine learning are proving that jobs previously thought to be the sole purview of humans can be done — as well or better — by machines.
Boston Consulting Group has predicted that by 2025 as much as a quarter of jobs currently available will be replaced by either smart software or robots. A study out of Oxford University also suggested that as much as 35 percent of existing jobs in the U.K. could be at risk of automation inside the next 20 years.
Take a look at these 10 professional jobs that are threatened by advances big data and machine learning:
Some aspects of a doctor’s job can now be done by computers. For example, surgeons already use automated robotic systems to aid with less invasive procedures. IBM’s Watson proved it can diagnose lung cancer from analyzing MRI scans much more reliably than real people. In addition, the UCSF Medical Center recently launched an automated, robotics-controlled pharmacy at two UCSF hospitals that automatically dispense prescriptions based on barcodes scanned by nurses. In fact, Johnson & Johnson has an FDA-approved device that can deliver low levels of anesthesia automatically — no anesthesiologist required.
Much of what insurance brokers and insurance underwriters do today can be done by computers using big data and machine learning. Formulas have been used for decades to determine how much insurance a person is qualified for and at what rate, but new tools will automate the decision-making process even more.
Programs already exist to help individuals design their own homes, making architectural skill and even design and color choices more automated. For now, most people are using the software mostly as a visualization tool, or to replace architects for very small projects. But as the sophistication of the programs improves, so will the need for human architects and designers diminish.
Much of what journalists do can now be automated using machine learning tools such as narrative science that creates natural language news stories from analyzing data. In fact, if you’ve read a financial earnings report in the past year or two, you’ve probably read an article or press release generated by a machine. The first places these programs will be used is in financial and sports reporting, which rely heavily on data and numbers, but other fields are not far behind. Services are already appearing that “scrape” content from news sites and “rewrite” it to avoid outright plagiarism but include the same content for websites.
- Financial industry
Algorithms can now analyze financial data and prepare accounts (as well as do tax returns) — without the need for accountants. Bank tellers have already been partially replaced by ATMs, but soon even higher level bankers, including loan officers, could be easily replaced by automated systems. Even governments are now using big data and machine learning to check tax returns and identify potential fraud in tax matters. We know that computers are already being used to make stock trades faster than humans ever could and they’re even used to predict how the market will react and make recommendations whether you should buy or sell.
The job of teachers will definitely change with the digitization. Studies have already shown that algorithms used to customize leaning to individual pupils based on their progress and understanding can be more effective than a human teacher. While this may be a boon to school districts desperate to find qualified individuals to teach, it may also eventually reduce the role of classroom teacher to that of proctor or babysitter — or eliminate it altogether.
- Human Resources
Human resources, headhunting and hiring is already being affected by data mining as algorithms take on the job of sorting through resumes to find the perfect candidates. Other jobs of human resources, including collecting and filing paperwork, advising employees about benefits, etc., can easily be automated.
- Marketing and Advertising
Marketing is all about that most human of skills, persuasion and manipulation. But even that is being successfully outsourced to computers. Persado, a natural language software firm, has put its computers to the task of writing compelling email subject lines for large retail organizations that can as much as double open rates. Companies are also experimenting with automated ad buying: instead of having people choose which magazines to place ads in and on which pages, the computers take care of it, using billions of data points for reference.
- Lawyers and Paralegals
In the discovery phase of a lawsuit, lawyers and paralegals can be required to sift through thousands, even tens of thousands of documents depending on the case. Now, sophisticated databases can use big data techniques including syntactic analysis and keyword recognition to accomplish the same tasks in much less time. In fact, it’s likely that a Watson-style machine learning system could be legally “trained” to review precedent and case history and even draft legal briefs — which has traditionally been the job of lower level law firm associates. But don’t think it’s only the lowly junior associates whose jobs are at risk: lawyers are well paid now to predict the outcome of major cases, but a statistical model created by researchers at Michigan State University and South Texas College of Law was able to predict the outcome of almost 71 percent of U.S. Supreme Court cases. That ability to predict outcomes is possibly the most valuable (and lucrative) service lawyers provide, and it was easily matched by a computer.
- Law Enforcement
Predictive policing is a hot-button topic. Many critics say that predictive policing is an infringement of civil liberties, but it’s not all as “Minority Report” as many people believe. In 2003, the same sorts of algorithms retailers like Wal-Mart use to predict demand for products was used to predict demand for police presence in New York City on New Year’s Eve, and the results were striking: 47 percent fewer random gunfire incidents, and a $15,000 savings in personnel costs during the 8-hour period. Better risk prediction could decrease the number of officers needed at any given time and for any given department.
The crux is: computers threaten more than low-skill jobs like factory workers, retail clerks, and waiters. As computers become exponentially more sophisticated, it naturally follows that they will be able to perform more sophisticated work. This will be a boon in many industries with increased accuracy and productivity. Any doctor would tell you that more accurate diagnostics are a good thing, and any lawyer would agree that faster, more comprehensive discovery is a benefit to the legal process.
The problem, however, lies in the fact that these technological revolutions might not create as many jobs as they eliminate. Certainly we will need more programmers, statisticians, engineers, data analysts and IT personnel to create and manage these sophisticated computers but it might be difficult to tell a factory line worker or taxi driver to shift gears and become a data analyst. How we fill the gaps when jobs are replaced will be the deciding factor as to whether all this automation is good for humanity or not.
This article was written by Bernard Marr from Forbes and was legally licensed through the NewsCred publisher network.
BNY Mellon is the corporate brand of The Bank of New York Mellon Corporation and may be used as a generic term to reference the corporation as a whole and/or its various subsidiaries generally. This material does not constitute a recommendation by BNY Mellon of any kind. The information herein is not intended to provide tax, legal, investment, accounting, financial or other professional advice on any matter, and should not be used or relied upon as such. The views expressed within this material are those of the contributors and not necessarily those of BNY Mellon. BNY Mellon has not independently verified the information contained in this material and makes no representation as to the accuracy, completeness, timeliness, merchantability or fitness for a specific purpose of the information provided in this material. BNY Mellon assumes no direct or consequential liability for any errors in or reliance upon this material.