Can AI prevent selfie accidents?
Where did you pose for your last selfie? Was it in front of the mirror showing off your outfit, before a well-known landmark in the city, or was it hanging over a cliff to share the view at your vacation destination? Did you fear for your life?
My last selfie was at Almedalen – see image below – after just finishing my final panel discussion.
Almedalen is a little bit like a Swedish “Davos”. It takes place in Visby, Gotland, every year in the first week of July. Politicians, organizations and experts come together to talk about the future they foresee in thousands of seminars, panels and debates. I was fortunate to join this year – it’s a great meet up – with many brilliant people and interesting ideas.
Is AI a threat or possibility?
During the week I took part in several panels. One was the Microsoft seminar on ‘Artificial Intelligence: Threat or possibility for humans and society’*. Apparently, it’s an engaging topic since we needed to turn people down due to lack of space in the hall.
And it was an engaging topic for me as well. As you might know, if you have read my previous blogpost, I am drawn to exploring the more questionable side effects that come with new digital technology. When technology is meant to aid and enrich our lives but also brings queries around ethics, privacy and security. So of course, I had to bring some of these examples to the panel. And I thought I could share one of them here with you.
Researching how AI can prevent accidents
Just before the Microsoft seminar, I read about a team of researchers that were exploring how AI can help consumers avoid accidents. And not just any accidents but the ‘Death by Selfie’ ones. Meaning where you aim to take a selfie in such an extreme place or in an advanced way that you actually end up killing yourself. Yes, it is known to happen. According to the article the researchers had gathered information on 127 cases. More than half had happened in India but the accidents also occurred in Pakistan, US and Russia.
The researchers were working on implementing AI functionality so that the camera could give the photographer information on whether the location (the place where you want to take your selfie) is dangerous. The idea is to help you stay alert and avoid an accident.
Using AI to help people avoid accidents is of course all well and good. But as you also may know, if you have stayed up-to-date with consumer research we have done, is that safety tools also can also make people more risk averse. This is the smart device safety paradox.
So what effect would a selfie-warning-AI have on the photographer? Some might take it on as a challenge, but most would probably listen and avoid taking the selfie. But then, as we become used to the AI warnings, what would happen if the AI missed alerting us at a dangerous place? Would we stop doing our own assessment, and just assume it was safe since the AI was quiet?
Asking the right questions about AI
It’s food for thought. New technology will have effects on our actual behavior and it is generally good to reflect what that might be before implementing. If we start using AI as a support system, how will it impact our critical thinking? And do we really need AI advice for everything?
Anyhow, the panel went really well, and different thoughts on risks and benefits of AI were brought up. In general, there was a positive outlook for the future.
The rest of the week continued with many interesting discussions for me and my colleagues attending the event. And after my last panel I went straight out and did what many of us do when traveling. I found the nicest view (from a cliff in Visby) and took a selfie to share.
I wish you all a great summer, but stay vigilant of high cliffs and train tracks. And when you take a selfie, please, be aware of your surroundings!
Interested in more on how AI is affecting our society? Check out a couple of my AI-related posts:
*[[Artificiell Intelligens: Hot eller möjlighet för människa och samhälle]]