Brzoska wants Meta to pay for the fraud. Expert: Imposing penalties is ineffective
Rafał Brzoska’s fight against digital giants fueled the public debate regarding their monopoly position and the responsibility due to us. However, in regulating the operation of websites, we are still doing too little and everything too late.
A few months ago, my mother received a phone call. My very nervous voice on the phone informed me that I had caused a road accident. Therefore, I urgently need money – about PLN 30,000. PLN for the victim’s treatment or for a bribe to a policeman; in her nervousness she didn’t remember this detail. Such a conversation could kill an octogenarian with heart problems.
Mom knows quite a lot about the types of telephone scams and knows the case histories that are publicized in the media as a warning. However, there is a key thing that distinguishes this situation from others for which it could have been prepared. It was my voice, faked, as he swore, perfectly.
It’s easy to believe. I have been working in journalism for 20 years and samples of my tone, timbre and intonation are easy to find on the Internet. All you need to do is feed them to some synthesizer, probably relatively inexpensive and easy to find for free, even on the so-called dark web, i.e. forbidden areas of the Internet. Similar synthesizers will also create, if necessary, an image or video – all thanks to the progress of mathematical models, which today we call artificial intelligence, or AI. Unlike ordinary fakes, i.e. fakes, we call them deepfakes by association with the concept of deep learning or deep machine learning.
How many fakes are there? This is probably a stupid question today. According to some estimates, between half and three-quarters of all Internet traffic is the activity of non-human entities, which we usually call bots. (The word robot itself was introduced in 1923 by Karel Čapek, inspired by his brother Josef, who coined it from the Czech robot – serfdom).
Chat GPT creator Sam Altman sometimes suggests that we are close to a world where AI tools look at our computer screens. Bots are supposed to browse the Internet for us, and they prepare the content themselves. And while it is easy to imagine a world in which the partial share of generic materials in what we read and watch every day is not so bad, the practice is very far from this idea.
