Risk-Selfies, a.k.a. Self-Nomination for the Darwin Award
While visiting the north rim of the Grand Canyon a few years back, I recall an “avid vacationer” climbing up onto a tall, narrow boulder no wider than his feet at the edge of what was maybe a thousand-foot drop. The wind was blowing a fair bit and I heard a few mumbled “what an idiot” comments… but he did make it. Apparently some people do not.
Death by misadventure-while-picture-taking is roughly equivalent in frequency to death by fireworks in the US, where explosives are a common feature of Independence Day celebrations and are readily available in many states. One does wonder whether we capture all of the data here; I would guess a lot of “death by fall” is hiding some of these.
As it turns out, there’s some truth in that guess. Reading the underlying paper, the top causes of death in these situation are heights, water, heights+water, and trains.
But our guesswork is not complete. Annotators hired by the researchers tagged many sample selfie photos involving vehicles as unsafe, but has not caused selfie deaths, at least ones mentioned in the media since 2014.
The researchers in this piece hope to build an algorithm using AI classifiers that might prevent the taking of selfies in dangerous settings. I predict failure in that approach — all that will accomplish is people taking even more risks to defeat the anti-danger software in their phones.