This AI Predicts How Children Get Older. Can It Protect Them?


Predicting how many years one just takes it for granted and looks very hard to be good, especially in the difficult early years of adolescence. And yet sorcerers, drug lords, and other illegal gatekeepers of this age group readily count that.

Their predictions are often misplaced. Now London-based London digital identity company Yoti believes that estimating the age of AI can predict a person’s age if they are between the ages of 6 and 60. For the first time, it says, they can accurately determine if children are at least 13 years old or older. to be.

Yoti’s visual effects can be as intriguing as Big Tech and the online experience more and more how children use their medications. However, proponents of privacy claim that analysis of people’s faces often leads to more stereotypes, often lawlessness, and bias.

Yoti says his year-old expertise, which he has developed over the past three years, has a limit of 2.79 years for all of his 45 years. For those under 25 years the error limit drops below 1.5 years. Over the next few weeks, it will test the bricks and mortar in five major UK markets. The company does not name supermarkets, but says several pornographic sites and unspecified games are also experimenting with technology. leave guests of a younger age. It adds that its age-old expertise is already being used by Yubo’s children’s playgrounds and health program Smash.

Point to the camera with the Yoti app on your face — it can work online on your phone, laptop, or tablet, or on your own — and the machine will measure your age. In several tests using the browser interface on my phone, the machine placed me neatly in sections 27-31 and 28-32. The company claims that even its customers do not keep the photos they take, and you do not have to register to use them. “I don’t understand. It does not guarantee the safety of any individual, “said Julie Dawson, director of regulatory and policy at Yoti.

Yoti customers can also use the limits in comparing age: for example, implementation limit of 25 when someone in the UK buys alcohol they must be over 18 legally. Anyone who can be identified as being below the limit can be asked to provide an ID. The system also informs its customers how confident they are in any comparisons.

The company has trained neural networks and “hundreds of thousands” of facial images, says Yoti cofounder and CEO Robin Tombs. It has often collected these faces, through its Yoti standalone program that allows people to verify their IDs with governments and other agencies by uploading official documents such as passports and driving licenses. When people put more information on the Yoti program, they have a better chance of getting out of the information that is used to teach Yoti AI. The company does not know what AI uses on the face to determine the age of people. “We have to be honest,” says Tombs, “we don’t really know if it’s a wrinkle or a rash or something. It has done so much for us and now it is doing very well. ”



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *