The announcement put many people on alert: OpenAI, the rising company behind ChatGPT and DALL-E, revealed that it is considering allowing users to create content not suitable for minors with artificial intelligence, including “erotic imagesextreme gore and profanity”… Is everything allowed when it comes to exploring the power of new technologies or are there limits to innovation? We find ourselves facing a discussion about the values of what we build.
Although OpenAI’s rules currently prohibit the creation of explicit sexual content, in an internal document that was known a few months ago It was revealed that the company is working on lifting those barriers. Joanne Jang, a product manager at the company, told US Public Radio that They hope to start a public discussion about whether erotic texts and nude images should always be banned.
“We want to make sure people have maximum control as long as they don’t violate the law or other people’s rights. This excludes, of course, enabling deepfakes,” said Jang.
Although when asked if these rules could create porn images, he answered that “that depends on your definition of pornography.” This revelation comes when there are already applications that “undress” people and that have generated a stir around the worldwith use in minors in different locations in the provinces of Buenos Aires, Chaco and San Juan.
The debate takes us to a discussion that is several decades old, but still generates controversy: does technology have values? Popular wisdom often repeats that artifacts are “neither good nor bad” but rather it all depends on how we use them. However, philosophical reflection from the humanities settled the debate a long time ago: When we talk about technology, there is no such thing as value neutrality.
Tim Berners-Lee, creator of the World Wide Web in 1989, is one of the greatest promoters of the need to incorporate ethical rules into the very design of the technology we are building: “As we design the system, we are designing society. Everything has to be presented as something that we think will be a good idea as a component of who we are.”
Thus, to evaluate whether a device or platform is good or bad, it is not enough to look at the technology itself, but we must also consider the people who are creating them and transmitting their values to them. As Erin Young, a researcher at the Alan Turing Institute, states: “Technology is not neutral. It is shaped by the people who build the technologies, shaped by their choices and their values. AI systems are not objective. “When we train it with biases, its results are biased.”
If we recognize that technology is loaded with values, OpenAI must assume that its own are at stake when designing its products. Just because you can make a platform to create synthetic pornography doesn’t mean you should have to or that it’s a good idea. Understanding the difference is a job for companies, states and users.