Future Tense newsletter: What algorithms can learn from a single photograph.

Future Tense Newsletter: What Algorithms Can Learn From a Single Photograph

Future Tense Newsletter: What Algorithms Can Learn From a Single Photograph

Future Tense
The Citizen's Guide to the Future
March 15 2017 1:30 PM

Future Tense Newsletter: What Algorithms Can Learn From a Single Photograph

103801608
A picture can be worth more than a thousand words.

Purestock/Thinkstock

Greetings, Future Tensers,

Maybe you should think twice before hitting “share” on that photo. As former Amazon chief scientist Andreas Weigend wrote this week, photo-analyzing software has advanced to the point where it can recognize faces, deduce place and time of day, speculate whether you’re in a fancy restaurant or gay bar, guess your emotional sentiments, or even copy your fingerprints. As these algorithms bring us closer to a post-privacy world, he argues, “we need to start thinking about how these images of us might be used to make decisions about us”—and how we might protect against algorithmic discrimination.

Advertisement

Engineers are also creating algorithms with the potential to predict something else significant about us—when we’ll die. But, says end-of-life care researcher Ravi Parikh, that may not be as unsettling as it seems. In their increasingly accurate prognoses, he explains, these mortality-prophesizing machines may actually give us more humanity.

Here are some other things we read between generating Texas oilmen aliases for our all climate change–related correspondences:

Whack hacking claims: Despite some fearmongering reports, the WikiLeaks documents detailing CIA hacking tools do not show that the spy agency has compromised secure messaging apps like Signal. Instead, it shows they found risky, expensive, hard-to-scale ways to hack the phones they run on, writes Yael Grauer. They didn’t “break Signal any more than looking at your phone over your shoulder breaks Signal,” one expert told Grauer.

Dumped, again: Trey Herr explains that though we don’t know who provided the CIA files to WikiLeaks last week, political rivals have taken notice of the damage that leaking their opponents’ espionage tools can do. Expect a lot more of these sorts of dumps in the future.

Advertisement

5 fast facts about Heavy.com: Will Oremus gives us the lowdown on Heavy.com, the site that’s been dominating your Google news search results, in the signature quintet style the site has come to be known for.

Events:

Could technology—from high-tech helmets to virtual training to real-time biometric data—make sports safer? And how will it change the state of play? Join Future Tense in Washington, D.C., on March 23 for drinks and conversation with those working to sideline injuries. RSVP to attend in person or watch online here.

Algorithms tell us what to read, where to go, and whom to date, but do we really understand them? Join ASU’s Ed Finn, author of the new book What Algorithms Want: Imagination in the Age of Computing, and the New Atlantis Christine Rosen in Washington on March 28 for a conversation about why we need to understand the systems that increasingly steer our lives.* RSVP to attend in person or watch online here.

Bon voyage, Boaty McBoatface,
Kirsten Berg
for Future Tense

*Correction, March 15, 2017: This post originally misstated the location of the Future Tense event about the book What Algorithms Want. It will be held in Washington, not New York.

Future Tense is a partnership of SlateNew America, and Arizona State University.

Kirsten Berg is an assistant editor for Future Tense.