Guns: Can technology help prevent shootings?
Could Uvalde be the catalyst that finally makes “smart guns” a reality? asked Jordan Winters and Ken Dilanian in NBCNews.com. Like something out of a James Bond film, smart guns “are designed to be fired only by an authorized user, employing fingerprint detection, Bluetooth links, and other technology that locks the gun to anyone else.” Two firms say they’re close to bringing the technology to market, although efforts in the past—including a 2002 New Jersey law that would have required ID features—have been quashed by gun rights lobbyists. The technology in all likelihood would not prevent a mass shooting, but with individual gun ownership on the rise, it could prevent the loss of lives in accidental shootings, like that of the 3-year-old who shot and killed himself in Florida when he found a loaded weapon in his home. “Firearms safety group Everytown for Gun Safety counted nearly 400 unintentional shootings by children last year alone, killing 163 people.”
America’s gun violence problem won’t be solved until it’s “harder for people to acquire and use guns,” said Casey Newton in his Substack newsletter but, barring that, we are stuck with monitoring social media. Unfortunately, tech platforms keep failing at that. The Uvalde shooter, Salvador Ramos, favored a social media app called Yubo that “allows users to broadcast themselves live to a small group of friends.” Yubo “says it uses machine learning” to flag concerning posts, but apparently “those alarms weren’t connected to effective enforcement mechanisms.” The gunman used Yubo, “to threaten rape and school shootings” in posts that were reported “dozens of times,” and nothing happened. In schools, tech companies are promoting algorithms that monitor students’ online activity, said Corin Faife in The Verge, but here again there’s not much evidence the approach works. The Uvalde school district contracted with one such social media monitoring tool, called Social Sentinel, “which claims to identify and alert schools to threats based on social media conversations.” The company monitors only public posts, and so missed private messages from the shooter that could have provided a warning.
Identifying potential killers by using computers to analyze vast amounts of data for subtle signs is another area of investigation, Jeremy Kahn in Fortune. One approach combines data about past violent incidents from schools with other data, such as gun purchase records. Another sifts through transcripts of sessions with psychiatrists. “I think it is possible that this kind of AI could help flag troubled individuals who are at risk of committing violence.” But in the end, we’ll be left with the same problem that afflicts most “predictive policing”: Just what do we do with the information that someone might carry out a shooting?