A new push for privacy [New York Times]
In the latest post-GDPR privacy push, Google’s CEO Sundar Pichai has penned an op-ed pitching privacy for all, not for the few. After a long list (TL:DR) of Google’s pro-privacy efforts, Pichai finally makes the real pitch, a push for federal privacy legislation in the U.S. over a state by state approach. Referencing how GDPR “raised the bar” for global privacy regulation, he asks for legislation head on, claiming it will help “ensure that privacy protections are available to more people around the world.” But, our savvy readers will know that streamlined legislation is simply better for business.
AI in blue: predicting police shooting before it happens [The Atlantic]
Since 2015, the US has been rolling out the use of body-cameras on police officers to record their actions on duty. One body-camera company is stirring controversy with its new tool. The company has developed an algorithm that can predict when a cop might be more ‘unstable’ or might resort more quickly to violence based on the body camera footage, internal police reports on officers’ behaviour and their environment on duty. It’s not an absolute science, but it can ‘score’ the risk of a police shooting. Are we prepared for the ramifications of predicting human behaviour activity using AI?
Sport is increasingly adopting AI. Sports fans can rest easy, it’s not replacing players (yet). AI is helping teams examine the performance of rivals and their own players alike, aiming to improve performance. After tennis, golf and basketball, now it’s football’s turn to benefit from AI. Semi-professional English football club Leatherhead FC has been using AI since last summer. The result? AI didn’t necessarily help the club finish higher up in the league, but it did improve player relationships as factual information could back constructive criticism. As the team’s assistant manager Martin McCarthy said, “it’s allowed us to build strong human relationships.”
Punny algorithms? [Wired]
AI can do a lot of cool things, but can it be funny? That’s what a Stanford researcher He He has tried to create. It turns out it’s pretty tough to get the neural networks that AI uses to learn to be flexible enough to successfully create puns. So, He and her team taught the algorithm the formula that humans (subconsciously) use to create puns: a surprise factor and homophones. The result? “The greyhound stopped to get a hare cut.” Still some work to be done…
In case you haven’t had enough…
AirPods are a tragedy [Vice]
It’s time to break up Facebook [New York Times]