TeaThere is a lot of concern these days about the dangers of human-AI relationships. Reports of suicide and self-harm due to interactions with chatbots have naturally made headlines. The phrase “AI psychosis” has been used to describe the plight of people who experience confusion, paranoia, or dissociation after talking to large language models (LLMs). Studies have shown that young people are increasingly embracing the idea of AI relationships, further increasing our collective anxiety; Half of teens chat with an AI companion at least a few times a month, with one in three having a conversation with an AI.Being more satisfied or complacent than real-life friends,
But we need to put the brakes on panic. danger Are There are real, but also potential, benefits. In fact, there is an argument to be made that – depending on what future scientific research reveals – AI relationships could actually be a boon to humanity.
Consider how ubiquitous non-human relationships have always been for our species. We have a long history of engaging in healthy interactions with non-humans, whether they’re pets, stuffed animals or beloved objects or machines – think about the person in your life who is completely obsessed with their car, even saying its name. In the case of pets, these are Real Relationships Insofar as our cats and dogs understand that they are in a relationship with us. But our one-sided, antisocial relationships with stuffed animals or cars are things we don’t even know exist. Only in the rarest of rare cases do these relationships develop into a pathological condition. For the most part, parasociality is normal and healthy.
And yet, there is some unrest about AI relationships. Because they are fluent language users, LLMs create the supernatural feeling that they have human-like thoughts, emotions, and intentions. They also generate sycophantic responses that reinforce our viewpoints, rarely challenging our thinking. This combination can easily lead people down the path of confusion. it is No Something similar happens when we interact with cats, dogs or inanimate objects. But the question remains: Even in cases where people are unable to overcome the illusion that AI are real people who actually care about us, is this always a problem?
Consider Loneliness: One in six people on this planet experiences This, and this is associated with a 26% increase in untimely deathsEquivalent to smoking 15 cigarettes a day. Research is emerging This suggests that AI companions are effective in reducing feelings of loneliness – and Not just by acting as a distractionbut themselves as a result of parasocial relations. For many people, AI chatbots are the only friendship option available to them, no matter how corny it may sound. Sangeeta Lal as journalist recently explained In a report on those turning to AI for collaboration, we shouldn’t be so quick to judge. “If you don’t understand why customers want this connection and why they’re looking for it and why they need it, then you’re lucky enough not to feel left out,” Lall said.
To be fair, it should be argued that the rise of new technology and social media has played a role in exacerbating the loneliness epidemic. This is why Mark Zuckerberg has faced criticism for his glowing endorsement of AI as a solution to a problem for which he may have been partially responsible for creating. But if the reality is that it helps, then it cannot be dismissed outright.
There is also research to show that AI can be used as an effective psychotherapy tool. in a studyPatients who chatted with an AI-powered therapy chatbot saw a 30% reduction in anxiety symptoms. Not as effective as human physicians, who achieved a 45% reduction, but still better than nothing. This utilitarian argument is worth considering; There are millions of people who are unable to access a physician for whatever reason. And in those cases, turning to AI is probably better than not asking for help at all.
But one study isn’t proof of anything. And there’s the rub. We are in the early stages of research on the potential benefits or harms of AI companionship. It’s easy to focus on the handful of studies that support our preconceived notions about the dangers or benefits of this technology.
It is in this research void that the real dangers of AI are revealed. The majority of organizations deploying AI companions are for-profit companies. And if there’s one thing we know about profit-seeking companies, it’s that they’re willing to avoid regulations and avoid evidence that could hurt their profits. They are encouraged to downplay the risks, cherry-pick the evidence, and report only the benefits.
The emergence of AI is not unlike the discovery of the analgesic properties of opium; If used by responsible parties with the goal of relieving pain and suffering, both AI and opioids can be legitimate tools for treatment. But if bad actors exploit its addictive qualities to enrich themselves, the result is either dependence or death.
I hope there is room for AI collaboration. But only when it is supported by strong science, and deployed by organizations that exist for the public good. AI must avoid the problem of sycophancy that leads the weak to confusion. This can only be achieved if they are explicitly trained to do so, even if it makes them less attractive as potential mates; This notion is anathema to companies that want you to pay a monthly subscription, without which you lose access to your “friend.” They should also be designed to help the user develop the social skills needed to connect with real humans in the real world.
The ultimate goal of AI companions should be to make themselves obsolete. No matter how useful they may be in addressing gaps in medical access or reducing loneliness, it will always be better to talk to a real human being.
Justin Gregg is a biologist and author of Humanish (Oneworld).
Further reading
Relying on code: living in the shadow of AI By Madhumita Murgia (Picador, £20)
The coming wave: AI, power and our future By Mustafa Suleiman (VINTAGE, £10.99)
Supremacy: AI, ChatGPT and the race that will change the world By Parmy Olsson (Macmillan, £10.99)
