Nearly half of all human jobs could be automated in the next two decades. If your job is repetitive and routine-based, there’s a significant chance you might soon be replaced by a machine – according to recent research on the future of human employment.
Bill Gates agrees that human career options will be significantly different in 20 years from now. „Technology over time will reduce demand for jobs, particularly at the lower end of skill set,“ Gates predicts.
But it’s not just unskilled workers, telemarketers and number crunchers whose careers are about to change. From news reporters to venture capitalists, here are five sophisticated and creative jobs that are already being carried out by algorithms.
1. VITAL, venture capitalist
Last week an algorithm-based piece of software was appointed the title of board member at Deep Knowledge Ventures, a venture capital fund based in Hong Kong.
VITAL (the Validating Investment Tool for Advancing Life Sciences) analyses economic trends and a variety of databases from life science companies. Based on its analysis the software makes recommendations to the board of members about venture capital strategies. According to the firm, it’s already helped Deep Knowledge Ventures make several investment decisions.
Although VITAL’s appointment to board member may just be a PR move, it is symbolic of how frequently algorithms are being used in venture capitalism. At the moment, the new board member’s input is still being closely supervised by his colleagues. However Aging Analytics, the research agency that developed the algorithm, says they wish to fully develop VITAL „such that it is capable of making autonomous investment decisions.“
2. Quakebot, robot journalist
Earlier this year, a computer algorithm called Quakebot made international headlines when it broke a news story about an earthquake in California. By speeding up the research process, the algorithm’s author was able to post the first report on the earthquake within three minutes.
Algorithmic decisions are typically based on a series of exact instructions that are used to process data, in this instance, information about a tremor.
Robojournalism has the potential to be employed in a variety of areas of reporting, and the LA times is also using algorithmic tools to report homicides. Rather than posing a threat to journalism, the algorithm’s author sees such the tool „makes everybody’s job more interesting“.
A recent US study showed that audiences could hardly differentiate between brief news stories written by humans and algorithms. Participants even agreed that the automated texts were objective, descriptive and trustworthy.
3. BORG, high frequency trader
First pioneered in 1989, BORG was a series of programming algorithms behind Automated Trading Desk, the world’s first high-frequency stock trading firm.
The predictive formulas were used to calculate profitable trades and execute the trades in just one second. High-frequency firms quickly boomed in the 90s, with as much as two thirds of all trades being conducted by speed traders.
Today, speed traders are making less money on a shrinking number of trades, making this a rare example of an area in which algorithm use is currently in decline.
4. Xerox’s HR algorithm, hiring boss
The average employer gets 144 applications per job opening. No recruiter can give each one of those applications a fair chance. Unless you’re not human.
When it comes to recruiting, several global firms are beginning to realise that you shouldn’t give a human a machines job. Several studies have proven that humans are worse at making good hiring decisions than simple algorithms.
To improve its long-term HR strategy, Xerox has developed a hiring algorithm that selects applicants for its 48,700 call-centre jobs. The software, which assesses answers to a series of multiple choice questions, is responsible for all recruitment in Xerox’s customer care department.
Algorithmic recruitment paired with extensive analysis of big data has helped Xerox reduce its turnover of employees by more than a fifth.
5. Google’s artificial brain, researcher of YouTube cat videos
Last year Google built a large-scale simulation of a neural network and let it run wild on YouTube. Distributed across 16,000 processor cores, the artificial brain analysed random clips of YouTube videos. Naturally, the first thing it does is use a deep learning algorithm to understand what a cat is. After a week of self-training, the network could identify cats with 75% accuracy.
Most significantly the algorithm proved the potential for computers to teach themselves without the help of humans. Although it is not yet been put to commercial use by Google, the software is predicted to used for facial and speech recognition.
Feature image: Vintage tin toy robot via Shutterstock / copyright: josefkubes