Breakdown
- A new AI will be used to write police reports based on body cam audio
- People are worried about how the police will abuse this system
- AI is also prone to writing racist and incorrect content
Police are using AI to generate police reports based on body cam audio, which sounds like a system that can easily be abused. Considering the stigma against police and how AI has been used to push certain racist agendas, this sounds like a recipe for disaster.
Forbes reports that this new AI system is called Draft One, which will supposedly help police write up their reports, ensuring they waste less time writing. While this might sound harmless on paper, the unpredictability of AI can make these reports seem worse than they are, especially since AI is no stranger to racism.
Axon, the company responsible for Draft One, claims that this AI won’t be going off the rails like a lot of people think it might. Principal AI product manager Noah Spitzer-Williams says that they “turned off the creativity” to ensure that nothing scandalous happens.
"The simplest way to think about it is that we have turned off the creativity," Spitzer-Williams told Forbes. "That dramatically reduces the number of hallucinations and mistakes... Everything that it's produced is just based on that transcript and that transcript alone."
It’s hard to trust that process since training AI can be difficult and we doubt that amount of effort will be put in a program designed to write reports. While there is a hope that policemen are going to check on these reports and make sure that they don’t report anything wrong, only time will tell.
Naturally, there are plenty of people who are hesitant about Draft One, given the recent reputation the police have had online. There’s been an extensive amount of racism reported, to the point that it inspired the Black Lives Matter movement, and it’s easy to see why people don’t trust the police.
Only time will tell if Draft One is going to be a successful AI program or another mess that’s going to make the lives of people harder. All we can do is hope that no one gets hurt or falsely accused because of these AI reports.