Why everyone suddenly doll? The latest direction of Amnesty International is more than harmful fun

If you pass social media recently, you may notice a lot of … dolls.

There are dolls All over x And Facebook feeding. Instagram? Dolls. Tijok? I have guessed that: dolls, in addition to Lessons On how to make dolls. Even there are dolls everywhere LinkedInIt can be said that the most dangerous and less organ of the gang.

You can call it AI or Barbie Box. Or if Barbie is not your thing, you can go with AI Action characters, Action Figure Pack or Action Action numbers. But however, you multiply it, the dolls apparently everywhere.

Although they have some similarities (boxes and packaging that mimic Barbie in Mattel, personality -dependent accessories, a plastic appearance smile), they are all different like people who publish them, except for one decisive and common feature: it is not real.

in New directionPeople use the AI ​​Gynecological tools such as ChatGPT to re -perceive themselves as dolls or work numbers, with accessories. It has been proven that she is very popular, not Only with influencers.

Celebritiesand Politicians and Main brands I have all jumped. Journalists who report direction Make Editions of themselves Holding microphones and cameras (although this journalist will not make you suffer from it). Users have made versions of any prominent character you can think of, from billionaire Elon Musk to actress and singer Ariana Grande.

According to Media Tech DrillingI already started on the LinkedIn social media, as it was common for the marketers Looking for sharing. As a result, many dolls you will see there are to promote a company or hustle. (He thinks, “Social media marketer doll“Or”Search engines doll (SEO). “))

But since then it has been leaked to other platforms, as it seems that everyone has enjoyed some fun to see if life in plastic is really great. However, it is not necessarily harmful, according to many artificial intelligence experts who spoke to CBC News.

“The wild West is there when it comes to the Public Prosecution,” said Anatoly Gruzd, a professor and director of research at the social media laboratory at the University of Toronto Metropolitan.

“Most political frameworks and legal frameworks have not fully innovated, leaving it related to artificial intelligence companies to determine how to use their personal data.”

Privacy concerns

Matthew Gozdiel, a professor of computing science at Alberta University, says the popularity of dolls is not surprising at all from a social point of view.

“This is the type of internet direction that we have since we had social media. It may have been customary for things like e -mail to be directed or testing as the results are shared,” he told CBC News.

But as with any direction of Amnesty International, there is there Some concerns On the use of data.

Artificial intelligence in general represents great special challenges to data. as Stanford University Institute for Human AI (Stanford Hai) notes, data privacy problems and the Internet are not new, but artificial intelligence “thirsty for data” to the point that it increases the size of the risk.

“If you are providing an online system with very personal data on you, like your face, function, or your favorite color, you must do this with an understanding that these data is not only useful in obtaining the immediate result – like a doll,” said Windy Wong, Professor of Political Science at the University of Columbia, who is studying human rights.

Wong explained that this data will be fed back to the system to help them create future answers.

In addition, there is concern that “bad actors” can use data embodied online to target people, as Stanford Hai notes. In March, for example, the competition office in Canada Warning against the rise of fraud linked to Amnesty International.

About two -thirds of the Canadians tried to use the Trucophachid intelligence tools at least once, according to New research by the social media laboratory at TMU. The report said that nearly half 1500 people, the researchers, who took samples of them did not have a sign or without understanding how these companies were collected or storing personal data.

Gruzd, with this laboratory, suggests caution when using these new applications. But if you decide to try, it suggests searching for an option to cancel the subscription to using your data for training or other third -party purposes under the settings.

“If there is no such option, you may want to reconsider the use of the application; otherwise, do not be surprised if doubts appear in unexpected contexts, such as online ads.”

The environmental and cultural influence of Amnesty International

Then there is an environmental impact. CBC Dogs and quarks I have already reported how Increased energy intelligence systems Technology with the ability to consume a lot of electricity like an entire country.

A A study from Cornell University It is claimed that training the GPT-3 language model from Openai in American databases in Microsoft can directly evaporate 700,000 liters of clean fresh water, for example. Goldman Sachs estimated that artificial intelligence An increase of 160 percent will pay for energy demand in the data center.

It takes a medium query ChatGPT around 10 times Energy more than Google search, according to some estimates.

until CEO of Openai Sam Altman She expressed concern about the popularity of images, and writing on X last month that she had to provide some borders temporarily while working to make them more efficient because her graphics processing units were “melting”.

Meanwhile, while the dolls created from our artificial intelligence take care of our social media summits, as well as a release that is circulated by the artists concerned about Reducing the value of their workUsing a retail sign #Starterpacknoai.

Fears have already raised about the last trend of Amnesty International, where users Pictures of themselves In the TOKYO Studio style GHIBLI – and I launched a discussion if it was He was stealing the work of human artists.

However, despite fears, Guzdial says these types of trends are positive – for artificial intelligence companies trying to develop their user bases. He said that these models are expensive to train and continue running, but if a sufficient number of people use them and depend on them, companies can increase subscription prices.

“This is why these types of trends are very good for these companies that are deep in red.”