Dear readers, dear readers:
We are a fascinating species. We tamed the majestic wolf and put him in winter coats to walk him around. We invented the internet and filled it with porn. We have within our reach a disruptive technology capable of generating moving digital images almost indistinguishable from reality… and we created a reality show on horns.
Spain will contribute a program presented by Raquel Sánchez Silva called Falso amor to Netflix’s global catalog on July 6th. The premise sounds familiar to us from similar programs: several couples are separated on two different paradise islands or villas and each group is surrounded by other very attractive human specimens and very interested in them, to see what happens. The innovation of the new contest comes from the use of deepfakes to “blur the line between truth and lies”, according to the promotional texts. Whoever wins gets 100,000 euros.
If human nature and some camera tricks were already enough to push the participants in this type of program to the limit, adding artificial intelligence to simulate that their partners are having sex with other people, put it in front of their eyes and play with the uncertainty as to whether or not what they are seeing is real sounds like psychological Guantanamo. The promotional video attests to the quality of the deepfakes, which superimpose the faces of the contestants on the bodies of doubles.
A tweeter has called the program “The AI ??of temptations” and I found it insurmountable. These days Netflix has heated up the promotion of the program with a deepfake video in which Sofía Suescun, former winner of Big Brother and regular television, appeared at her house with a man who is not her partner.
That Netflix dedicates itself to these things is scary, I don’t know if more or less than the fact that we are capable of creating this type of video for anyone else, or that others can create ours. Is it ethically acceptable to make porn for your own consumption with other people’s photos? Is it an invasion of the privacy of others or the technological evolution of a simple fantasy? Wired magazine asks, which recalls the case of a youtuber who mistakenly showed a tab in which deepfakes of other images appeared during a live show. two youtubers friends of hers.
The fact that the victims of these montages tend to be young or famous women can make us suspect what the correct answer is. Earlier this month we published in La Vanguardia an article by Susana Pérez Soler dealing with the subject in depth and explaining what to do if someone uses our image to create artificial porn without consent… while European regulation arrives -China has already limited it by law- . Last week we reported in this newsletter that pedophile material in which children’s faces are superimposed on adult bodies is increasing.
Of course, the issue of consent is key, and the objective of deepfakes is not only pornographic, criminal, misinforming (politicians are also regular protagonists, for example) or television. “I think it’s not a question of the technology, it’s how you use it,” Harrison Ford has stated in a very sensible way during the promotion of Indiana Jones and the dial of fate about the use of sophisticated techniques of creation of images and AI to rejuvenate you. .
“This is 40-year-old Harrison Ford, and that’s why he looks so good,” he said. “I have been working for LucasFilm for most of my adult life. Every frame of those movies we’ve made together can be analyzed with – here we go again – artificial intelligence. And they can find the right angle, the right light so that they are my mouth, my eyes, my face, fused together. It’s not a Photoshop retouch or anything like that. It doesn’t look like that. It is real”.
Summing up this week’s question: An artificial Harrison Ford in the movies? Good. Artificial horns on Netflix? Wrong. Am I falling into a Manichean thought? Sure, but one grew up in the comfort of Spielberg’s good guys and bad guys and would like to recapture the feeling for a couple of hours this weekend at the theater. The more willing we are to believe, the less sophistication is necessary in deception.
What else has happened this week
We continue to meet heroes of artificial intelligence. Joseph is a computer scientist who has managed to combine three jobs at the same time by teleworking and has tripled his salary to 300,000 euros per year. His companies don’t know, of course. When meetings coincide with him, he removes the camera and attends to each one for a while.
If there is a skull and a book in the lap of a saint, we are before a vanitas, an allusion to the transience of life. The Prado Museum and the Barcelona Supercomputing Center (BSC) are collaborating on the FrAI Angelico project, which trains AI models to recognize the symbology and content of the Prado paintings.
Xataka has published some very practical topics on the use of AI tools. For example, how to use the Playground image editor, how to use the Midjourney function that recreates a photo beyond its limits or how to use Chat GPT as a translator, to interact with a PDF or to play games. Another interesting practical text is from The New York Times in English: how to turn ChatGPT into your personal coach.
This desktop robot that uses AI to animate and accompany its owner while working is the saddest thing I’ve seen in a long time. It’s called Letianpai and it goes on sale at the end of July in China for about 270 euros.
The Vatican has published its own manual on the ethics of Artificial Intelligence. It can be downloaded here in English.
Microsoft president Brad Smith has said at an event in Brussels that artificial intelligence can be “a powerful weapon” and that “we have to work together to develop defensive AI technologies that create a shield capable of resisting and defeating the actions of any criminal on the planet”. The truth is that the more vehement the speech by companies interested in regulation, the more suspicious I find it.
Need a cheat sheet on how apocalyptic the experts on the subject are? It is here, and it turns out that they are much more so than they are, more likely to warn of the dangers present.
YouTube will soon implement automatic dubbing of its videos into other languages, and by 2024 this dubbing can be done with the voice of its author.
Google says its DeepMind will dwarf ChatGPT.
Wired magazine profiles some of the most prominent women working in AI.
Technology advances, but where are we going to put the data centers? Where are we going to get the electricity and water they need to function? In The New York Times, in English.
An interesting take on the modern pace of -and from- a powerful industry: the last Marc Jacobs show lasted only three minutes, the models wore their hair like Daryl Hannah in Blade Runner and the press release was written by ChatGPT. At Harpers Bazaar.
OpenAI is opening its first non-US office in London. There is also DeepMind, Anthropic… and a prime minister very interested in AI. In The Guardian, in English.
A list of movies on the subject to watch on long summer afternoons. In The Guardian, in English.
The internet continues to fill up with junk at a good rate. The Verge, in English.
Three other English-dense topics to pay attention to: while legislating, the EU uses sophisticated AI systems for border control (Le Monde). The use of facial recognition in stores in the United Kingdom raises many questions (NYT). United States, China and the geopolitics of AI chips (The Washington Post).
IAnxiety Levels This Week: On hold until we see the new Indiana Jones movie.
