Artificial intelligence toys could end up in

Artificial intelligence toys could end up in

Artificial intelligence toys could become smarter in the next 15 years than parents buy them for their children, according to a market scenario. But let that be the smallest problem!

“Elf on the Shelf”, Santa’s undercover agent, whom the duplicitous parents move to different places in the house to “spy” on their children and have the blackmail to be good all year round, could become a reality. Only the place of Santa’s elf army, encamped between the ice at the North Pole, will be taken by the multinational corporations in the toy industry.

Barbie, the ugly beauty

The prefix “smart” seems to stick to anything these days. We have smart phones and TVs, but we also have seemingly smart pens or toasters. Many times, however, this does not mean that the technology in question is really better, but rather that it comes at a higher price. A smart toy can learn, adjust how it interacts with the user, can react to external stimuli, and behave according to programmed patterns.

One of the first attempts was “My Friend Cayla”, a line of Genesis Toys dolls that could interact and talk to their little playmates. To do this, the puppets used voice recognition technology and searched the internet for answers to children’s questions, giving them a real-time reply. Launched in 2014, they have been named the Toy of the Year by the London Toy Industry Association, quickly entering the top ten sales of European retailers. Unfortunately, the dolls also recorded everything that was said about them and proved to be relatively easy to hack. The German authorities have described them as a risk to the safety of their families, recommending that their parents destroy them. Starting in 2017, the company stopped selling them.

Then came Hello Barbie, from Mattel, a doll that used artificial intelligence (AI) to “know” the child, then developing a specific communication profile, adapted to his personality. “It’s scary,” The Washington Post said. “It’s not just scary. It is also a threat to the safety of children, “added The Daily Beast. Once the child spoke to her, the doll recorded the conversation and transmitted it to the servers of ToyTalk, a San Francisco company through which Mattel breathed life into Hello Barbie as well. Speech recognition software converted audio content to text, and other artificial intelligence software selected certain keywords to suggest to the puppets an answer from a base of over 8,000 lines imagined by a team of writers. Everything in less than a second!

Not to forget, Hello Barbie had built in the cloud a database about children’s preferences and habits, a selection of keywords extracted from children’s discussions being channeled to a basis for analyzing consumer trends. She was also withdrawn from the market, following an extensive protest campaign under the slogan #HellNoBarbie.

In the meantime, technology has advanced significantly, and experts believe that toys animated by artificial intelligence algorithms will become a key component of how children learn and develop. “Without a doubt, the world in which these children will live will be a world enhanced by artificial intelligence, and this is extremely important to society, whether we like it or not,” said Richard Gottlieb, chief executive officer (CEO) of Global Toy Experts, a toy consulting firm. If we look at the numbers, the segment of artificial intelligence toys is growing rapidly and could reach 18 billion dollars by the end of next year.

At the moment, there are a variety of smart toys and kid-friendly AI gadgets on the market, such as the cute robot Anki Cosmo, the robotic-training puppy Aibo or even the children’s edition of Amazon Echo Dot. Basically, there has never been a better time to enter the world of (inter) connected toys. Although most of them may play a beneficial role in children’s learning, they will continue to (re) present significant risks as long as they are not designed with the confidentiality and security of user data in mind.

Dangerous connections

Properly designed and supervised, smart toys for children could become an incredible resource for their development, says Kay Firth-Butterfield, head of artificial intelligence and machine learning at the World Economic Forum. However, if we do not take the proper precautions, they could become the surveillance tools used to record every movement and every word of the children, along with everything that is done and said around them, she adds.

If you do a Google search for “smart toy safety”, the results may seem quite worrying. The volume of articles about how easy it is to hack a toy that uses Bluetooth or Wi-Fi connections could disappoint any parent. But why would anyone want to hijack a child’s toy?

CloudPets is a plush toy that allows family and friends to send messages to a child, messages played through a built-in speaker. It comes in the form of a dog, a bunny, a cat, or a bear, and as the Witch has already shown, it’s pretty simple to take control of it and get it to play other (and other types of) voice messages. The engineers of the platform broke the “kitten” version and then used it to do more online shopping, by voice command to an Amazon Echo device near the toy. As far as they can see, they were able to connect to CloudPets’ insecure Bluetooth connection even outside the building where the test took place, directly from the street.

“Such toys typically contain sensors, microphones, cameras, data storage components, and other multimedia capabilities – including speech recognition and GPS options,” the FBI said in a warning to parents, noting that they can be reprogrammed to function as 24/7 permanent surveillance devices, or to record video / audio to children, without the adults around them noticing.

But another kind of 24/7 surveillance already exists. And it’s done quite often even with parental consent. The child-oriented advertising market is a multi-billion dollar industry, and marketing companies are building their target audience profile from an early age to better sell their products.

“Advertising technology collects a significant amount of personal data about children during their online time. On average, we talk about over 72 million data points about a child before he or she reaches the age of 13, ”said Dylan Collins, CEO of Super Awesome – a company that strives to create a safe internet for children by developing technological solutions that stop online data collection and tracking.

When technology companies violate their privacy and privacy in order to collect data for marketing purposes, children are at risk, according to a UNICEF report, according to which advertising aimed at children can jeopardize their healthy development. Digital advertising is increasingly targeting children and can have long-term effects, such as childhood obesity, self-harm or the use of tobacco, marijuana and alcohol, according to the same source.

The real imagination

However, the (increasingly) growing popularity of AI toys also raises some ethical questions. Much of a child’s development is his or her ability to engage in creative play, in which imagination is the most important toy he or she owns. This is also why many child development experts prefer so-called “open” toys – such as Lego bricks or dolls – to the detriment of rigidly programmed AI. In the hands of a child, even a cardboard box can be a fairy castle, a pirate ship or a rocket to the moon. And for psychologists who study children’s play, the big concern is not that artificial intelligence could make them imagine exaggerated things.

On the contrary, they fear that their imagination may suffer. “Imaginary friends do not impose limits or constraints. Children often change their age, appearance, gender, priorities, or interests. In the case of a toy like Hello Barbie, its personality is limited from the start by programming, as well as the interaction. It can be a lot of fun, but it’s certainly less creative and interactive than playing with someone / something that can be anything and everything you want to be, “said Tracy Gleason, a psychology professor at Wellesley College (Massachusetts, USA).

Other psychologists warn that AI toys, especially humanoid robots, could have a profound impact on how children will perceive and build interpersonal relationships. To see how they react to life-giving technologies, roboticists at the Massachusetts Institute of Technology (MIT) let several children play with two humanoid robots. They were not able to talk to their little playmates, but they interacted with them through eye contact, gestures, and facial expressions. In the end, most children were convinced that robots could listen, feel, care, and be able to make friends. And that, even after the research team showed them how it works and even let them control it.

As a toy that can befriend a child has a commercial potential that is hard to ignore, it is to be expected that artificial intelligence will be more sympathetic, empathetic and friendly. “It simply came to our notice then. And for some children, artificial friends may start to replace the real ones. If you have someone who listens to you every time and you can talk to him all the time, why try to make friends ?! Family and friends can annoy, provoke, or disturb us, while an AI toy will always be nice to us, ”warns Noel Sharkey, a professor at the University of Sheffield, England, a pioneer of robotics.

For his part, Peter Kahn Jr., a psychologist at the University of Washington in Seattle, USA, points out that this kind of friendship could lead to what he calls a relationship of domination, in which the child is not just the one asking for everything. time, but also the one that is rewarded every time, without feeling any responsibility towards the robot. And such a balance of power, he adds, is unhealthy for his moral and emotional development. Because the abuse of power is just a step away. What is no longer a child’s play.

This article appeared in issue 141 of . magazine.

PHOTO: Getty