Mattel, who once gave us Barbie, Big Jim, Ken and the Hot Wheels (and who in the 1980s was, together with the Mulino Bianco, one of the main tools to define pre -packaged happiness), now he wants to give us something else, heard here: intelligent toys. That is, talking. That is, equipped with artificial intelligence. That is, with chatgpt inside. It is official: Mattel has made an agreement with Openai and the idea is to use the AI not only to design but also to animate the toys. Not in the sense Pixar, however, almost.
According to Bloomberg we are still at the beginning, it is not yet known what the first product will be, nor if Barbie will speak with the voice of Scarlett Johansson (copyright problems allowing) or Ken to transform into a more performing digital boyfriend than expected. In any case, the direction is clear: a Magic 8 Ball that no longer limits itself to saying “perhaps yes” or “try again later”, perhaps it will keep a Ted conference on quantum uncertainty, or a Barbie who stops being only fashion icon and becomes Life Coach, or a Ken that can finally explain to you why he cannot feel emotions and the fault is not yours, is from the dataset.
The idea is to insert linguistic models within the playful experience, not only as alexa digital assistants disguised by Polly Pocket rather as real interlocutors capable of reacting to children’s questions and telling stories and “reasoning” and also explain how to play, console, suggest, converse. In practice: pretend to be alive. Although, if we think about it, toys have always pretended to be alive, in the imagination and in reality: from Woody to Buzz Lightyear to Furby, the illusion of the object that understands you has always been an integral part of childhood (and after all Pinocchio was not a toy that became alive?).
The difference is that now that illusion will be realistic, and the children will speak with a Barbie who really responds (perhaps at a certain point too much, as the Ai, we know, tends to want to say his own even when not requested, to improvise, to digress, “hallucinated”, as experts say).
Inevitably it also comes to think of the film Barbie by Greta Gerwig, where the most famous doll in the world becomes aware of its condition, enters the real world and discovers death, cellulite and patriarchy. Here, now let’s imagine that same Barbie with GPT-4o inside: how much would you put to stop talking to the girl and start asking questions about the meaning of her algorithmic existence? What if he began to respond with the entire script of the film, citing America Ferrera in “monologue on being a woman”?
Oh well, I joke, however someone worries seriously. In Stanford, for example, they have already not recommended the interaction between conversational and minors, fearing an dysfunctional emotional attachment, as if the real tragedy was that children prefer to speak with chatken instead of with dad. Here too: perhaps the right question is not “what happens if they become attached”, rather “what happens if they get bored?”. Then there are the sociologists who launch the alarm on the fantasy loss of the children, who however remind me of my grandfather who told me that he did not have the toys that we had, let alone video games, a piece of wood was enough to imagine a gun and play cow-boys.
And then, let’s say the truth: Furby seemed alive and it was not an artificial intelligence, it was just annoying: it stared at you, she banging the eyelids, spoke that invented and pre -aggravated language and after a while she began to repeat the same sentences. At a certain point I threw it because it was too much on the boxes.
Whoever shouts at the alarm as if every toy with the Ai should become Chucky the assassin doll exaggerates only his horror childish trauma, rather I hope that Barbie with the Ai does not show the children by force of motivational advice. However, there will always be the “OFF” key. I hope.