Your stomach might’ve dropped reading the headlines this week with a stark warning from an AI researcher over artificial intelligence (AI) and the future of the human race.
For those who missed it, researcher Jacob Coxon said he’d resigned from Anthropic, an American artificial intelligence research and product company, warning that “the people building AI earnestly believe that it could kill us all by the end of the decade”.
“They are racing straight to self-improving superintelligence and gambling with our lives,” he added.
After a summer of climate change-fuelled heatwaves (and the anxiety that has created), some of us are now trying to navigate another looming existential threat – one that is hard to understand or comprehend.
Trying to shed some light on the situation, Jacob Coxon told CNN how, while the current risk from AI to humans is low, at the rate the technology is progressing, it could end up causing “extreme havoc”.
“For example, hacking critical infrastructure, building extinction-level bioweapons, there’s a lot of ways that the AI could actuate itself in the world,” he said.
Following Coxon’s X post, Evan Hubinger, alignment science lead at Anthropic, said his former colleague’s assessment was valid, adding that he believes there is an over 10% chance the technology could wipe out all humans within the next decade.
Hubinger caveated in a follow-up post that the risk from currently available AI models is low, but added: “What I am worried about is superintelligence arising from recursive self-improvement.”
In response, an Anthropic spokesperson told Wired: “We have always been transparent that AI will bring both enormous benefits and unprecedented risks.
“This work is also why we believe the world would benefit from the industry adopting a lawful, verifiable way to work together to pace how we release powerful models.”
This sort of information is deeply concerning to hear and will likely leave a lot of people feeling anxious.
Therapists are already noticing a rise in AI anxiety
Annabelle Hird, a counsellor and member of Counselling Directory, says it’s something people are already speaking to her about.
“At the moment I am hearing about AI in my practice at least once a day,” she told HuffPost UK. “Concerns range from issues involving AI in relationships – for example partners using AI to support their side in disagreements – to worries about job security.”
Counselling Directory member Ruby Mitchell is also seeing worries crop up more in clinic. In in the last 6-12 months in particular, she’s noticed a real increase: “It’s cropping up in my clinical work weekly or fortnightly at least.”
Psychologically, she says the concerns are “familiar territory” for therapists as they focus on “uncertainty, or a lack of control, and anxiety about catastrophic events”.
“Clinically, it feels similar to issues such as climate change – clients may say they feel something enormous could be looming, but also feel the power to change things lies in the hands of a small number of people,” she added.
BACP senior accredited psychotherapist, Debbie Keenan, expects AI-related concerns to start becoming more prominent in the therapy room. “AI is becoming part of everyday life – naturally, people will start to have concerns about their employment, privacy, safety etc and what the future holds for them,” she said.
“Sometimes this comes from fear of losing control or not being able to tolerate uncertainty. Life can be uncertain at the best of times, least along with this looming threat.”
Olga Pankova via Getty Images
How to tackle AI anxiety
1. Seek information from credible news sources
First things first, think about where you’re getting your information from.
“I would encourage people not to allow frightening headlines to become predictions about their future,” said Keenan. “Look at the facts and seek information from credible news outlets.”
If you’re feeling especially anxious, try to limit news intake and steer clear of social media, as “constantly checking for updates can heighten anxiety”.
2. Ground yourself
“When we are frightened, we are out of our ‘Window of Tolerance’ and dysregulated. Noticing what is around you, getting outside, [and] maintaining normal routines is essential,” said Keenan.
“Our attention is wherever our minds are – bring your attention back to the present moment, rather than catastrophing about the future. Focus on what you can control,” she added.
“It is okay to be concerned without letting that concern take over your life. Stay informed, but stay grounded in what is factually known, rather than becoming consumed by what might happen.”
3. Trust in the experts
Hird suggests also placing a degree of trust in those who are meant to keep us safe, including those working in AI.
“I am old enough to remember all the anxiety leading up to the year 2000 around the millennium bug, or the YK2 bug as it was otherwise known. Many reported all kinds of potential disasters related to how the date would impact technology – [but] when the turn of the millennium came there were actually very few minor disruptions that occurred.”
She continued: “One of the scariest things about AI is that, to so many of us, it is an unknown entity. We are having to trust that those in the know are making the correct choices – but we have very little understanding of what those choices might be.
“One thing we can be certain of is that thoughts shared publicly are almost certainly overhyped and likely driven by some sort of agenda. We can also be certain that those working in AI are the best placed to keep us safe.”
4. Focus on what you can control
Look at the areas in your life where you do have control, suggests Keenan, as this can help give us agency when other elements feel uncertain.
“We can take responsibility for ensuring we read trusted sources of information about the topic, explore the consequences of our use and make choices that align with our values when using available products,” she added.
Ultimately, Mitchell notes that telling people not to worry is “unrealistic and perhaps even unhelpful” given there are legitimate uncertainties around AI.
She likens it to if a client came to her with a physical symptom and started going down a rabbit hole of health anxiety. “It’s not really my job to say ‘you’re going to be fine’,” she said. “As an anxiety specialist, I do know that where anxiety becomes problematic is when the mind starts treating uncertainty as if it were imminent catastrophe.
“Human beings are not especially good at living with an unknown future. We tend to fill the gaps, and anxious minds generally fill them with the worst available scenario – it’s kind of our prehistoric brain trying to protect itself against the worst outcome.”
She noted one useful distinction is between being informed and being saturated with information, “so you can perhaps follow developments in AI without reading every warning that appears online”.
And while a certain amount of worry and anxiety are part of what it means to be human, the expert warned if it begins impacting your functioning, sleep, relationships and day-to-day life, it might be time to speak to a professional about it.
Mitchell ended: “Human beings, with the right support, can adapt (and historically have adapted) to many frightening situations … And we will continue to do that here.”
