Joseph Crawford of the University of Tasmania and Michael Cowling of RMIT University discuss the dangers of emotional dependence on advanced technology.
As of November 2022, generative Artificial intelligence (AI) has been rewriting billions of emails, student assignments, reports and breakup text messages. Along with this common usage, we have seen an increase chatbots as AI “friends”.
Users can design and dress (or undress) their own your best AI friendusually for a fee. These social chatbots provide “good conversations” and a relationship that seems perfect when it loses the same vulnerability and genuine sense of togetherness that we gain from relationships with other people.
Proponents argue that for those suffering from social disconnection, chatbots are available as unlimited, supportive friends who are not judgmental. And there is some evidence chatbots can alleviate symptoms of loneliness, chronic disconnection and blocked existence.
In Australia eSafety Commissioner admits this: “AI friends are always available – which can be fun for those young people who have trouble forming face-to-face relationships.”
But the the dangers of changing human relationships chatbots are just beginning to be understood.
Research we did in 2023 showed university students using early chatbots to get academic support and deal with unintended consequences, such as loneliness. Currently, the accidents The downsides of social chatbots – such as dependency, defriending people, social withdrawal or creating awkward relationships – seem to outweigh the benefits.
In recent weeks, China has set new rules for AI partners and chatbots, or “anthropomorphic interactive services” as they are. called in the law. Along with guaranteeing that chatbots must protect national secrets, the emotional use of these tools is limited, and expected to follow social and ethical standards.
These are some of the stricter controls on public use of AI delivered anywhere in the world. So what can we learn?
China’s social AI law
Significant damage from the concurrent use of AI is occurring, including in China. An 18-year-old child he spent 7,000 yuan (A$1,400) on her AI boyfriend’s custom jewelry over their four months of “dating”.
A lonely family 84 years old in China’s Hubei province he found out that he would be surprised by the AI character, spending about 10 hours a day watching videos with him and other characters, spending thousands of yuan on the service and writing love letters to him by hand.
China’s new laws establish clear provisions for how they expect friendly chatbots to interact with people.
The measures expect public AI bot developers to ensure that content does not:
- they threaten national security or national honor
- they violate core socialist principles
- encourage self-harm or suicide
- excessive attention to emotional dependence or addiction
- provide ‘partners’ or ‘relatives’ for children
- provide inappropriate content to children.
In addition, users who appear to be overly dependent on the service should receive prominent reminders of what they are talking about with the chatbot. And AI companies need to step in when users seem too stressed.
The risk of a ‘loneliness spiral’
As we’ve shown, there are concerns – and studies so far suggest this too – that it’s unhealthy for young people to rely solely on AI as a trusted friend or therapist.
Lonely people facing social anxiety can share problematic use of AI conversations to reduce feelings of loneliness and negative feedback.
We suggest that this reliance can lead to an AI-induced ‘loneliness spiral’. When people feel lonely, they become very serious to feelings of social threat and disruptive emotion regulation.
Therefore, they turn to a low-risk AI ‘friend’ for support, instead of trained or friendly humans. However, this usually doesn’t fix the initial problem of loneliness, because chatbots can make it worse.
The restrictions China has placed on AI partners could help break this cycle and re-prioritize human interaction. It also scales with mental math – would it be more satisfying to install an AI-enabled app, or to make an appointment with my friend?
Australian studies
Loneliness and a low sense of belonging remain a major challenge in Australia, at least one in six to be lonelywith some evidence which suggests higher prices.
While China has been ramping up its AI partners, Australia has been at the forefront of youth use of social media. eSafety has also developed codes for to restrict age-related material on the Internet such as pornography, extreme violence, self-harm, suicide and disordered eating. But they don’t talk about the social harms of AI companions, such as emotional dependence or removing human interaction.
Prime Minister Anthony Albanese too he did not speak social harm from AI chatbots in his big talk on AI last week.
We suggest that Australia’s age-restricted protections should extend to the unintended harm of social interactions with social, analytical and functional AIs.
The Australian government needs to examine how these true friends communicate with the Australian people. And we must prepare to reduce the unintended harm to society before these social interactions become widespread and mainstream.
Australia must prioritize improving social connections – in workplaces, manufacturing techniques, education curricula and the structure of our societies.
The despotism of chatbots may not reach a democratic nation. But we can – and should – prioritize developing social skills to counter an unhealthy overreliance on AI.
And Australia should use legal protections to reduce public harm, expanding on the work eSafety is already doing with social media.
Joseph Crawford again Michael Cowling
Joseph Crawford is an award-winning management, leadership and higher education scholar based in the Office of the Vice Chancellor for Education University of Tasmania. He is interested in leaders and how their behavior shapes the world around them. This includes organizational leaders from various fields including teachers as leaders and student leaders.
Professor Michael Cowling is an award-winning technology educator and consultant with a keen interest in mixed reality education, computer science education and information technology in society. He is currently a professor and director of the Hub for Apple Platform Innovation at the School of Computing Technologies in RMIT University and is the founder of The Create Lab, which focuses on collaborative research and engagement in technology and education.
Don’t miss out on the information you need to succeed. Sign up for Daily BriefSilicon Republic’s digest of must-know sci-tech news.
