Concerns over AI are causing heated debates as each aspect of our everyday lives continues to saturate every aspect of us. Recent studies have provided some insight into how AI can be used to make a more equitable and less marginalizing environment for those in the LGBTQ community.
These issues have been preparing. According to anthropologist and gender studies specialist Mary L., it was widely believed that AI may “always fail Gay people” by 2021. Gray responded, “While LGBTQ+ areas continue to evolve, it is the very reality that we grow and develop our names, and our society, which means AI may struggle to understand LGBTQ people,” Forbes said.
People who identify with these organizations “continually push what is probable,” according to Gray as a result of the development of LGBTQ+ rights. However, this may have a positive impact on LGTBQ+ people, but it can have an impact on AI, which claims to do its best work when it has everything stable..
What are the problems with the Gay area and AI?
Part of the problem is that AI “[builds] decision-making models by looking at existing or ‘staple’ decisions,” which it then tries to replicate, Gray said to Forbes. So by design, AI “excludes and pushes to the margins anything that doesn’t have a robust example,” said Gray. When trying to involve people who identify with a, this can be hazardous.
And while many LGBTQ+ individuals are employed in the AI industry, there is a “divergence between the gay individuals interested in artificial intelligence and how the same group of individuals is growing their business,” he added. The system “universally responded with typical depictions of gay society” when AI was asked by Wired reporters to create an image of a queer man,” according to the store.
These AI-generated benefits “frequently presented a stereotypical, white version of gay life,” said Designed. The Artificial system that Wired used, Midjourney, created outcomes where “lesbian people are shown with nostril rings and harsh emotions. Gay people are all stylish dressers with criminal abs. Basic pictures of are hypersexualized, with clothing outfits and cleavage-focused camera points.”
And when it comes to the LGBTQ+ group, AI is doing more than just generating pictures. According to one group of European experts, they have “developed an AI unit that can tell whether you’re gay or straight.” According to these researchers, their AI model can “differentiate between homosexual and heterosexual men with an accuracy rate of 83%” by “examining subjects’ electrical brain activity.”
Reviewers say that this design, like Midjourney, oversimplifies gender and identity. According to Qinlan Shen, a research professor at Oracle Labs, “Bisexual and infertile individuals exist but are ‘simplified apart’ by the Swedish study to make their empirical setup workable.” This implies that AI systems can and will be used as a means of security and repression in places where LGBT+ expression is hated, Shen said.
How can these problems be addressed?
There is no distinct answer. A “human being evidently does a way better job because they draw out details in conversation that a system can’t,” he said. This is more apparent when AI “creates” thoughts, images, or even the dreaded “social media content.” “Artificial intelligence gets trained to be as sanitized as possible — great as a resource for major commercialization,” said the Chronicle.
There are “plenty of important applications of this technology, many of which have been in use for centuries now,” said the Chronicle. But as a “replacement for artistic expression, there’s just no real explanation for AI that holds up.”
But, there are LGBTQ+ individuals working to improve AI. Sam Altman, one of the most well-known figures in the field, is one of them. “I’m sure no one would choose to work in this field because they’re homosexual because artificial intelligence and technology is a “field with a lot of key representation and a lot of opportunity,” said Altman, who is explicitly gay. Despite Altman’s arguments, the fears over picture in AI remain.
