Science, Weeklies

Man versus machine

Society may be down to the final sprint in the race against the machine.  As technology progresses, more and more jobs that were once held by humans are being filled by machines.  As machines become smarter and smarter, this trend can only be expected to continue into fields that are traditionally thought to be safe from automation.

“There are a ton of pluses to have machines do factory-type jobs because they can do jobs that are dangerous for humans, and they are more efficient, faster and more precise,” says Samantha Monarch, a sophomore in the College of Arts and Sciences. “As far as actually taking jobs from humans, I think they will just require workers with a higher skill set. People won’t be needed to do the jobs the machines are doing, but companies will need workers who are qualified to repair, program and oversee the work of machines.”

“The major change that has happened over the last 20-30 years is that tasks that can be modeled algorithmically have been replaced by technology. These are not necessarily unskilled and the tasks that are not readily replaced are not necessarily skilled,” describes Boston University economics professor Kevin Lang via e-mail. “Cleaning a room is currently too difficult for us to address algorithmically. Bookkeeping is relatively straightforward. As a result we have lost a lot of bookkeeping jobs.”

The limits of what can be considered a “book-keeping job” have been pushed in recent years. A computer program built by a Chicago-area start-up called Narrative Science can automatically write news briefs.  Articles heavy in information and low in personality can be easily modeled.  For example, the program is able to take statistics from a sports game and turn them into a game summary of the game.

“Machines taking the place of humans in today’s job market is silly. In this economy, people will do anything to get a job, no matter what. Also, there are so many people that come to this country looking to work,” says Katarina Kozomora, a junior in the Sargeant College of Rehabilitation Sciences.

The Narrative Science website describes their product, “Our proprietary artificial intelligence platform produces reports, articles, summaries and more that are automatically created from structured data sources. With amazing speed and quality, narratives are created in multiple formats, including long-form articles, headlines, Tweets and industry reports.”

A computer scientist at the University of Washington, Oren Etzioni, told the New York Times, “The quality of the narrative produced was quite good,” and that it is very human like.

This program is beginning to be used by companies and newspapers as a way to provide low cost news briefs and to analyze other data.



This is not the only area that machines are gaining ground.  Earlier this year, the world was shown Watson, IBM’s Jeopardy playing super computer.  In a game against the best Jeopardy players in the past ten years, Watson won by a healthy margin. Besides games, the computer also has aspirations of real world applications.

The IBM website states, “Jeopardy! The IBM Challenge poses a specific question with very real business implications: Can a system be designed that applies advanced data management and analytics to natural language in order to uncover a single, reliable insight — in a fraction of a second?”

“[Watson’s] real test will be applying the underlying data management and analytics technology to different industries,” the IBM website reads.

“I found Watson to be really scary. If machines are taking the place of humans in jobs, then how am I supposed to compete with that?” asks Mia Greenberg, a senior in the College of Comunication. “I guess factual recall isn’t as valuable as analytical skills, but I’m sure Watson isn’t far from that either.

However, these new and exciting computer technologies may not be putting people out of work. The Narrative Science website insists that the technology is meant to aid and enrich the work of journalists, not replace them. For instance, their technology may be used to provide a quick information filled template for an article that involves large amounts of data or be used to cover areas that normally do not get any media attention, such as youth sporting events.

Watson’s application goals share a similar philosophy with Narrative Science’s work.  Well Point Inc, a large health insurance provider, has licensed Watson to help choose treatments and diagnose patients from its database of 34 million members.

“This very much fits into the sweet spot of what we envisioned for the applications of Watson,” said Manoj Saxena, general manager of an IBM division told the AP.

Because these new products are not looking to replace humans, such technologies may not be responsible for the recent rise in unemployment.




“None of this is particularly related to the recent rise in unemployment. There was a lot of technological change in the 1990s, but the unemployment rate was low,” writes Lang. “It is much more closely related to the rise in inequality in the United States, both because many of the jobs that have been shifted to technology were mid-level jobs like bookkeeper, typist, bank teller and because technology has facilitated outsourcing of low-skill jobs.”

Two economists, Erik Brynjolfsson and Andrew P. McAfee of MIT, have a more bleak outlook on the job market in their new book, “Race Against the Machine.”

In their book, the authors warned that automation is accelerating and this increase in automation is putting people out of jobs.  The professors write in the book, “This last repository of jobs is shrinking — fewer of us in the future may have white-collar business process jobs — and we have a problem.”

They claim that the rate of automation is increasing and entering new fields, which means more and more humans will be replaced in the coming years than ever before.

Unlike Watson and the work that Narrative Science is doing, there are some technologies being developed that are very obviously seeking to replace humans and fit with Brynjolfsson and McAfee’s outlook.




Google has a group of automated Toyota Priuses that have logged over 190,000 miles on every kind of road.  This technology seeks to replace drivers and has the potential to make an entire industry of truck drivers obsolete.  The car uses an assortment of sensors and radars to build a 3D map of the area surrounding itself and then uses this map to avoid collisions and stay on track. Right now, Google has a human driver in every car for legal and safety reasons, but the cars have required very little intervention and the technology is being improved everyday.

“As a COM major, it’s a little intimidating that machines are being programmed to do things like write news stories,” says COM senior Rose Milgrom. “However, cars that drive themselves sound intriguing. I imagine it’s really expensive though, and I question how safe it really is. Roads are unpredictable! It’s not as easy as applying an algorithm.”

As a result of the automation trend, there are some worries that if computers continue to replace humans in the job market, output could become higher than demand.  This would create a situation where society has become very productive, but there is no one to buy the products that are being produced.

Lang dismisses this notion though, “Despite fears among lay people that growing productivity would create a situation where demand could not keep up with output, experience shows that we can have dramatic economic growth without ‘over-producing.’ Some people believed that the Great Depression of the 1930s was due to such over-production. We produced a great deal more in 1999 without high unemployment.”

As always, the future is uncertain.  While the threat of a job market being dominated by machines exists, we must remember that this fear has existed for hundreds of years. New technology always creates jobs along with the jobs it makes obsolete.

One Comment

  1. on . March 7th, 2012 at . 12:17 pm