Monthly Archives: November 2017

Scientists can now figure out detailed, accurate neighborhood demographics using Google Street View photos

By Christopher Ingraham

A Google street view car on display at the Google I/O developers conference at Moscone West Convention Center in San Francisco in 2013. (John G. Mabanglo/European Pressphoto Agency)

A team of computer scientists has derived accurate, neighborhood-level estimates of the racial, economic and political characteristics of 200 U.S. cities using an unlikely data source — Google Street View images of people’s cars.

Published this week in the Proceedings of the National Academy of Sciences, the report details how the scientists extracted 50 million photographs of street scenes captured by Google’s Street View cars in 2013 and 2014. They then trained a computer algorithm to identify the make, model and year of 22 million automobiles appearing in neighborhoods in those images, parked outside homes or driving down the street.

Street View scene with parked vehicles in Brooklyn (Google)

The vehicles seen in Street View images are often small or blurry, making precise identification a challenge. So the researchers had human experts identify a small subsample of the vehicles and compare those to the results churned out by their algorithm. They that the algorithm correctly identified whether a vehicle was U.S.- or foreign-made roughly 88 percent of the time, got the manufacturer right 66 percent of the time and nailed the exact model 52 percent of the time.

While far from perfect, the sheer size of the vehicle database means those numbers are still useful for real-world statistical applications, like drawing connections between vehicle preferences and demographic data. The 22 million vehicles in the database comprise roughly 8 percent of all vehicles in the United States. By comparison, the U.S. Census Bureau’s massive American Community Survey reaches only about 1.6 percent of American households each year, while the typical …read more

Read more here::

Robots could soon replace nearly a third of the U.S. workforce

By Danielle Paquette

Over the next 13 years, the rising tide of automation will force as many as 70 million workers in the United States to find another way to make money, a new study from the global consultancy McKinsey predicts.

That means nearly a third of the American workforce could face the need to pick up new skills or enter different fields in the near future, said the report’s co-author, Michael Chui, a partner at the McKinsey Global Institute who studies business and economics.

“We believe that everyone will need to do retraining over time,” he said.

The shift could displace people at every stage of their career, Chui said.

By 2030, the researchers estimated, the demand for office support workers in the U.S. will drop by 20 percent. That includes secretaries, paralegals and anyone in charge of administrative tasks.

During the same period, the need for people doing “predictable physical work” — construction equipment installation and repair, dishwashing and food preparation, for example — will fall by 30 percent.

Other advanced economies, such as Germany and Japan, will see at least a third of their workforce similarly disrupted, the report concludes.

China’s share will be smaller (12 percent), since more employers there will still find it cheaper to employ humans.

Machines can increasingly perform tasks that people have long handled. They scan Tylenol and lip balm at the drugstore. They build pickup trucks. They take your grilled cheese order at Panera.

Technology could replace up to 375 million employees worldwide by 2030, the McKinsey authors estimate.

The jobs most at risk involve repetitive tasks. About half the duties workers handle globally could be automated, according to the report, though less than 5 percent of occupations could be entirely taken over by computers.

Caretakers, psychologists, artists, writers — anyone who relies on empathy or creativity at work — can expect to …read more

Read more here::