Texts, images and videos produced with the help of generative artificial intelligence proliferate and, apparently, could have a effect on memory and memories. Our memory does not work as a digital archive. It is much more similar to a reconstruction process: every time we remember an event, the brain reworks it, colors it of emotions and fills any voids with more or less plausible details. It is precisely from this fragility that the so -called originates mandela effectthat is, the belief of remembering something that has never happened in reality or that has happened, but in a different way. The advent of the images and videos generated by the Amplifies the chances of running into false memories. Seeing more and several times a false but convincing content can make us believe that it belongs to our past, or even that it is part of a memory shared by many. Let’s see more in detail What is the Mandela effect And above all, What is the role of the AI in provoking false memories.
THE’mandela effect – that phenomenon for which a large number of people shares the memory of a fact that never happened – owes its name to the researcher Fiona Broomewhich in 2009 he noticed a curious phenomenon: many people were convinced that Nelson Mandela is died in prison in the 80swhen instead it was freed in 1990 and died only in 2013. Since then, examples of collectively incorrect memories have emerged in various areas of pop culture. Some examples? Many remember the Pikachuthe famous yellow Pokémon, with a black strip on its end (absolutely not present); still others remember theOmino del Monopoly with the monocol (In this case too, the detail is absolutely absent).
Psychology tried to explain why the mind is so inclined to this kind of distortion. Apparently an important part of the process is in thehippocampusa brain region involved both in memory and in the imagination. According to what was stated by prof. Aileen Oeberstof the University of Potsdam “This already suggests some important consequences for false memories»Adding that”If people imagine something repeatedly, they tend to believe at a certain point that they have really lived it and that it is basically a memory».
The border between remembering and imagining, therefore, is much more subtle than we think: if we repeat to ourselves an image or event that has never happened, with time we could convince ourselves that it belongs to our experience. In addition, each mnemonic call is not a faithful reproduction, but a sort of reconstruction, susceptible to external errors and influences.
The dr. Christian Jarretta cognitive neuroscientist, director of Psychein an article for the BBC expressed an interesting opinion regarding the phenomenon of false memories due to the spread of AI:
In psychology, a “monitoring of sources” describes the processes involved in identifying the origin of our memories. The picture records the source of a memory as information. It is codified in memory together with other aspects of what we have experienced. The “tag” that indicates the origin of the memory, however, can disappear easily, even if other aspects of the memory remain. In this way, the videos generated by the AI run the risk of confusing themselves with the events of the real world in our mind. And this is a problem that could worsen with the improvement of the videos generated by the AI.
With tools such as Dall-e, Sora (both of Openai), Nano Banana and Veo 3 (both from Google) today it is possible to create images and videos in seconds that perfectly imitate the style of an artist or that depict events never happened. Tools that allow you to produce deepfake (videos manipulated thanks to the AI in which one face is replaced with another) increasingly advanced and realistic, which in fact generate scenes that seem real, but which have never happened in reality.
The combination of vulnerability of memory and AI ability to produce convincing content places an important challenge before us: Learn to live with tools that amplify the phenomenon of false memories. Understanding how to mitigate the problem is not easy at all and at the moment. Certainly, an excellent starting point is to surround yourself with reliable sources, confronting competent people regarding the topics of our interest and maintaining a critical approach towards what we see online.
