Apr 13, 2017

Computers can be biased, just like us

Computers that learn words from texts written by humans capture their meaning but also our biases, a new study shows.

Why it matters: Machine learning is being eyed to sift through resumes in an effort to reduce discrimination in hiring, analyze loan applications and to predict criminal behavior while reducing racial profiling. The unintended biases found in artificial intelligence raise ethical questions about whether and how to deploy the technology without reinforcing stereotypes. (See Exhibit A, the racist Microsoft bot.)

How it works: The researchers created a test for how closely the AI associates different words and uncovered gender and racial biases similar to those of humans that are well-known from psychological studies. They found that European-American names were more closely associated with pleasant words (honest, gentle, happy) whereas unpleasant words (divorce, filth, jail) were more likely to be attributed to African-American names. Young people were considered pleasant, old people were not. They then looked at gender bias and found the AI associated women more so than men with family and the arts than with mathematics.

Thought bubble: Context provides bias but also meaning. How much bias can be removed before that meaning is lost?

The study authors don't recommend untraining the machine because of the risk of removing crucial knowledge about the world. "Artificial intelligence learns biases but it needs the awareness not to make prejudiced decisions. Since machines do possess self-awareness the way humans do, a human in the loop can help machines make ethical decisions," says Princeton's Aylin Caliskan.

Go deeper

Trump weighs quarantine of New York, New Jersey and Connecticut

Data: The Center for Systems Science and Engineering at Johns Hopkins; Map: Andrew Witherspoon/Axios

President Trump said on Saturday he is considering a "short term" quarantine of New York, New Jersey and parts of Connecticut — states that have already taken steps to quarantine residents and promote social distancing.

The big picture: With 112,000 people infected, the U.S. has the most COVID-19 cases in the world, exceeding China and Italy, per data from Johns Hopkins. A second wave of American cities, including Boston, Detroit, New Orleans and Philadelphia, are reporting influxes of cases.

Go deeperArrowUpdated 10 mins ago - Health

Q&A: Minimizing your coronavirus risk

Illustration: Aïda Amer/Axios

As the coronavirus continues to spread throughout the U.S., Axios is answering readers' questions about the pandemic — how it spreads, who's at risk, and what you can do to stay safe.

What's new: This week, we answer five questions on smokers' vulnerability, food safety, visiting older parents, hair cut needs, and rural vs. urban impact.

The other coronavirus test we need

Illustration: Sarah Grillo/Axios

Researchers are racing to develop tests that detect whether someone may have developed immunity to the coronavirus, which could help society return to normal faster.

Why it matters: These tests could help people know if they are able to go back to work, as well as aid researchers in tracking the scale and death rate of the disease — key data for current and future pandemic policies.

Go deeperArrow38 mins ago - Health