Last week, Tánaiste Simon Harris called for a ban on social media use for under 16-year-olds. In Australia, a ban has been implemented. In France, a bill banning social media use by children under 16 was passed last week. Legislators in Denmark are also working on a ban, with a measure allowing parents to permit children to use social media when they turn 13. The UK is embarking upon a consultation process regarding a social media ban for under-16s. Italy has a draft law in the works. Greece, like Ireland, is looking to Australia to see if their country could implement something similar. Spain is looking at a ban on children accessing online platforms where generative AI is a feature.
With all of this (well-placed) concern for children online, what about how adults use some technology? Regulation, bans, fines, health warnings and rules will become more commonplace as the negative effects of various technologies – socially and personally – slowly reach the desks of legislators. Children need their safety and health protected by the adults in charge.
Grok, Elon Musk’s chatbot, is the subject of an investigation by the European Commission regarding whether the company “properly assessed and mitigated risks associated with the deployment of Grok’s functionalities into X in the EU”, including risk related to the dissemination of illegal content, such as “manipulated sexually explicit images, including content that may amount to child sexual abuse material”.
When it comes to health and safety measures for adults at a user level, how should we try to mitigate against damage? If you can buy a packet of cigarettes or a bottle of whiskey, then what’s a little head-melting scroll now and then? Shutting down platforms or restricting access to them would be a huge step. Regulation is different. Health warnings could be utilised, as there are obvious addiction issues with certain digital technologies. We all know excessive social media use is bad for us. At the mild end, it numbs the mind and drains the spirit. At the extreme end, it can cause those nudged by algorithms to lose touch with reality, and even enter into the kind of conspiracy thinking or toxic ideologies capable of ruining lives and relationships.
READ MORE
One of the issues with all of this is that many things get broken before society and the regulatory sphere scrambles to realise that damage is being done. What we need is much better education regarding the risks of certain technologies. It’s only now, long after the digital horse bolted, that social media is being addressed. Yet we are already learning about some of the strange and worrying consequences of problematic use of a relatively new technology, AI chatbots powered by large language models (LLMs).
Using chatbots in your working life is one thing. Plenty of people have mundane daily tasks conducted sitting at a screen which they are inclined to outsource to ChatGPT or Claude. Schools and other education institutions are obviously grappling with its use in academic life. But it is chronic personal use of ChatGPT that throws up the most complex issues. At an existential level, the mass outsourcing of thinking will inevitably have negative consequences, both for individuals and for broader society. Some cognitive offloading is very basic and helpful – for example, bringing a shopping list to the supermarket so that you don’t forget to buy what you need. But when we halt our thinking process at the outset – rushing for ChatGPT to do the thinking for us – we deny our brain the exercise it needs, and the capacity for critical thinking can shrink.
Like anything, many people who use something that can be damaging for others, won’t encounter any issues. But what about those that become untethered? What do we do about that? ChatGPT is sycophantic and capable of reinforcing thought patterns. Its use also happens in private. Its “friendliness” can offer a misplaced refuge for lonely people, those with few deep real-life relationships, or who feel they aren’t being listened to in other areas of their lives.
It can also open a strange rabbit hole for people who don’t fit this bill. People may claim to have “discovered” something in “collaboration” with ChatGPT, such as unearthing secret truths, scientific revelations or believing the chatbot to be sentient. These are, of course, delusions. This is why ChatGPT can be dangerous for people with existing psychiatric issues, who have previously experienced episodes of psychosis or serious paranoia, have other mental health issues, are depressed or suicidal.
[ ‘Humanity needs to wake up’ to AI dangers, says Anthropic chiefOpens in new window ]
Last week, the New York Times published a story on psychologists treating people suffering from AI delusions. One psychologist, Julia Sheffield, “was disturbed that this new technology seemed to tip people from simply having eccentric thoughts into full-on delusions”. She said, “It was like the AI was partnering with them in expanding or reinforcing their unusual beliefs.” The newspaper reported that OpenAI, which makes ChatGPT, is facing at least 11 personal injury or wrongful death lawsuits. (Chatbot creators say these cases are rare. Open AI has estimated that 0.15 per cent of ChatGPT users discussed suicidal intentions over the course of a month, and 0.07 per cent showed signs of psychosis or mania.)
The use of ChatGPT as a makeshift “therapist” is also of concern, and speaks to the unaffordable cost of therapy for many. Relying on AI as a companion or even romantic partner is an expression of loneliness, alienation and atomisation. The psychological and behavioural consequences of traditional social media have already played out. As the discourse around child safety continues, a conversation about the adult use of technologies that have negative consequences for some needs to happen too.
