{"id":1029,"date":"2019-12-13T04:27:43","date_gmt":"2019-12-13T04:27:43","guid":{"rendered":"https:\/\/www.danielparente.net\/en\/2019\/12\/13\/a-new-study-of-self-driving-technology-finds-a-truly-stunning-flaw-really-its-all-too-human\/"},"modified":"2019-12-13T04:27:43","modified_gmt":"2019-12-13T04:27:43","slug":"a-new-study-of-self-driving-technology-finds-a-truly-stunning-flaw-really-its-all-too-human","status":"publish","type":"post","link":"https:\/\/www.danielparente.net\/en\/2019\/12\/13\/a-new-study-of-self-driving-technology-finds-a-truly-stunning-flaw-really-its-all-too-human\/","title":{"rendered":"A New Study of Self-Driving Technology Finds a Truly Stunning Flaw. (Really, It’s All Too Human)"},"content":{"rendered":"

[ad_1]
\n
<\/p>\n

\n
\n

The promise of artificial intelligence is twofold: It’s supposed to free us from life’s mundane routines, and also, ultimately, to offer slivers of improvement over what humans can do.<\/p>\n<\/div>\n

\n

Repetitive motion? Boring thought processes? Check, AI will handle them.<\/p>\n<\/div>\n

\n

Bigger numbers, more calculations, faster reactions than humans? That’s part of the promise.<\/p>\n<\/div>\n

\n

But it turns out that AI-empowered applications can amplify our less attractive qualities, rather than remedy them. It makes sense since we humans are the ones doing the empowering.<\/p>\n<\/div>\n

\n

Case in point: a new study out of the Georgia Institute of Technology<\/a> that says the self-driving cars appear to be better at not hitting and killing people with lighter skin tones. Put differently, the algorithm involved arguably incorporates the programmers’ implicit racial bias.<\/p>\n<\/div>\n

\n

It’s an early study, and not yet peer-reviewed, so take it with a grain of salt for now. But it found that on average, AI models were about 5 percent less likely to identify (and thus, avoid hitting) dark-skinned people as compared to lighter skinned people.<\/p>\n<\/div>\n

\n

The theory behind this is that programmers are unintentionally including their biases — even selection biases that they don’t recognize — when they create algorithms.<\/p>\n<\/div>\n

\n

As a Vox article reporting on this suggested: “Since algorithmic systems ‘learn’ from the examples they’re fed, if they don’t get enough examples of, say, black women during the learning stage, they’ll have a harder time recognizing them when deployed.”<\/p>\n<\/div>\n

\n

As a Vox<\/em><\/a> article reporting on this suggested: “Since algorithmic systems ‘learn’ from the examples they’re fed, if they don’t get enough examples of, say, black women during the learning stage, they’ll have a harder time recognizing them when deployed.”<\/p>\n<\/div>\n

\n

We’ve seen this kind of story before. Amazon<\/a> accidentally created a hiring algorithm that was trained used 10 years of applicants’ resumes. Unfortunately, they were overwhelmingly male candidates, which taught the algorithm to value women’s applications less.<\/p>\n<\/div>\n

\n

So, it’s not that self-driving cars are racist, or that automated hiring systems are sexist. Instead, let’s consider this a good, early opportunity to assess some of the real challenges with using AI that we thought would solve themselves.\u00a0<\/p>\n<\/div>\n

\n

Here’s what else I’m reading today:<\/p>\n<\/div>\n<\/div>\n

The opinions expressed here by Inc.com columnists are their own, not those of Inc.com.<\/p>\n