A post on Twitter with images of “reconstructions of historical figures”, including the alleged face of Guanche’s mummy made with artificial intelligence, caused a lot of controversy. So much so, that a week later the author of the tweet retweeted it, only this time making it clear that it was an “unreliable” reconstruction. Midjourney, one of the platforms capable of creating an image from a sentence, has used Midjourney, and some archaeologists and historians have raised a red flag to warn of the dangers they pose to cultural knowledge. They say that heritage reconstruction made with artificial intelligence (AI) creates Fake Historically, it has reduced scholarly work to a click away, reinforcing stereotypes and compromising copyright.
How to perform medically unreliable facial reconstructions using # Midjourney.
I’ll be detailing the easy process and all of the prompts used in my most controversial thread of 2023 (so far) so you can experiment too.
Here it goes 🧵👇 pic.twitter.com/YfV45BB9HU
– Javi Lopez ⛩️ (javilopen) January 15, 2023
Pablo Aparicio, owner of a company that performs 3D or 2D virtual reconstructions of heritage, explains that his work combines historical and anthropological knowledge, with various documentation techniques, to carry out authoritative reconstructions of architectural assets, works of art and archaeological sites. A single project can take many months of effort, even if multiple people work on it.
To reconstruct a castle, for example, Aparicio explains, laser scanning or digital photogrammetry is first done from the remains of the building itself. With a “perfect engineering base”, the architectural premise of the building is put forward with computer software such as Blender or other 3D software. We started to develop hypotheses in collaboration with architects and archaeologists who chart all the steps for us. We show them the progress and they tell us, for example, if the walls will be higher or lower, if they look different, where a roof is needed. They give us details on how to proceed, ”confirms the founder of this company, 3D Stoa – Patrimonio y Tecnología. “It is a job that must be very precise and concrete. We always have to take care of the modifications proposed by those who asked us for them, ”he continues.
The first problem with generative AI reconstructions, he argues, is a lack of specificity. As far as you ask summoned [el comando] Well written that designs a castle for you in a certain way, it is impossible to give it all the details so that it can do something specific for you, ”adds Aparicio. The images are very eye-catching and attractive and work well on social networks. , but with a sense of credibility that leads to confusion.” They present an image that looks real or looks like it could be real. And herein lies the biggest problem. It helps a lot to transfer files Fake “It’s historical,” says the archaeologist and historian, “It doesn’t have anything scientific, but that might sound right.”
The reason is the nature of this type of technology, which uses unknown data banks as input, which is often anything traded on the Internet. The end result is a mixture of everything on the internet. “Artificial intelligence is pulling out what is most prevalent. They will not represent the Colosseum rebuilt or complete, as it was in ancient times, but they will represent it in ruins, as it is today, which is also not entirely true, “says Aparicio.
Alberto Venegas, Professor of History and PhD from the University of Murcia, asserts that the main function of these tools is recombination, which is why they reaffirm stereotypes and perpetuate false impressions of the past. “It doesn’t even remotely resemble the past, it’s just a massive amalgamation of all the images, regardless of source, intent, or who made them. It’s a continuation of myths or images that are already commonplaces we’re used to seeing in the media,” he asserts.
To exemplify this, Venegas cites the “fantastic representation” of the city of Paris during the French Revolution. Through the lens of a GoPro camera which also circulated successfully on the networks. In his opinion, “any historian” will see that the Paris in these images is not the Paris of the Revolution, but the Paris of the Second Empire and the Third Republic. However, in the collective imagination, there are “Sound of Paris” elements, such as rocky roofs, chimneys, large doors, stone buildings, and grand avenues. “Paris in the eighteenth century was not like that, it was later when a series of reforms were carried out in the city. In the same way, the current French flag comes out, which differs from that time, ”he adds.
“AI has captured everything that could be sensible, and it makes us think of the French Revolution, which we saw on TV, video games, series and movies. I call it a media past and an aesthetic memory, which really makes us assume that time, but it is not Really,” Venegas explains over the phone.
Of course, the AI can be trained and its algorithm refined to improve the results obtained. This is the case for the work of Bas Kurstin, who tried to teach an artificial intelligence to replicate Rembrandt’s works until he could imitate them almost completely. pic.twitter.com/7ke4NlWjyj
– Alberto Venegas (@Albertoxvenegas) November 7, 2022
Will it be possible to use artificial intelligence to reliably reconstruct heritage? Venegas suggests yes. “If they give them the right data, they can get close to expert reconstructions,” he says. To illustrate, the professor recalls a project carried out by Design Executive Director Bas Corsten with Delft University of Technology and the Rembrandtweis Museum in the Netherlands, which collected 170,000 visual sources of the work of the Dutch painter Rembrandt. By processing this batch of images, the machine was able to produce an almost perfect artwork, following the artist’s style and brush.
Reconstruction tasty It also stirs controversy in the copyright field, appropriating the work of thousands of photographers, illustrators, designers and artists. While the debate about whether plagiarism is still open, the tools have the advantage of not being regulated.
Although Venegas and Aparicio have criticized heritage reconstruction using this type of tool, both agree that they are useful in small stages of the traditional process, such as choosing surface textures and other finishes. Also, in educational training. “At school, at the institute or at the university, they can use them to carry out reflection activities, criticism or to discuss different pictures, from different pasts, to draw conclusions. They are positive tools for the educational process, ”concludes Venegas.
You can follow country technology in Facebook And Twitter Or sign up here to receive The weekly newsletter.