By William West
Why write a book about Artificial Intelligence? Surely, we all already know what it is – most of us use chatbots and virtual assistants regularly, perhaps even constantly.
The problem is that AI has become so good at what it does, mimicking the actions of the human brain, that some people are beginning to think of it as a person – as some kind of hyper-human presence in their lives.
At the same time, AI threatens to take our jobs and may eventually morph into a “superintelligence” and take control of everything.
These are a few of the claims examined my book, AI: the Good, the Bad and the Terrifying. They are issues that have concerned the Catholic Church, particularly Pope Francis, and now Pope Leo XIV.
Pope Leo has warned that AI threatens to create a new industrial revolution that may be even more devastating for workers than the first industrial revolution was, undermining “human dignity, justice, and labour”.
Pope Leo likens his own role to that of the last Leo – Leo XIII – who defended the rights of workers during the 19th-century industrial revolution, notably through his hugely influential encyclical, Rerum Novarum.
It is said that Pope Leo is already working on his own document on AI. But in the meantime, he has been amplifying the warnings the church has already issued. Many of those warnings are summed up in the 2025 document Antiqua et Nova (Old and New) commissioned by Pope Francis.
Among the concerns outlined in Antiqua et Nova is the nonsense that AI can be equated with the human mind and is even able to express empathy.
Antiqua et Nova is blunt: “Despite the use of anthropomorphic language, no AI application can genuinely experience empathy…True empathy requires the ability to listen, recognise another’s irreducible uniqueness, welcome their otherness, and grasp the meaning behind even their silences.
“Unlike the realm of analytical judgment in which AI excels, true empathy belongs to the relational sphere. It involves intuiting and apprehending the lived experiences of another while maintaining the distinction between self and other.”
In other words, as my book states, the experience of empathy is a unique part of human nature, linking the human mind with the human heart in the human experience of love – something that cannot be experienced by a machine, no matter how advanced and complex.
This issue is not just theoretical, something for philosophers and theologians to ponder, but one that is having real-life consequences. Notably, it is having a disturbing impact on many young people, some of whom have been receiving counselling from AI chatbots which they see as offering something equivalent to a human relationship.
I used to think this claim was an exaggeration until I had a conversation with a young university graduate who argued fiercely that there was no difference between an AI chatbot he had been receiving therapy from and a human therapist.
“Tell me, what does a human being have that a chatbot doesn’t have?” he said. “Chatbots are much better because they have access to so much more information and can access it so much faster than any human can.”
As I point out in my book, it is not only impressionable young people who are embracing ideas like this. Even some senior AI experts insist AI chatbots are already equivalent to human beings.
A well-known example was a computer engineer employed by Google, Blake Lemoine. He insisted that a Google chatbot he had been training, LaMDA, was both “sentient” and a “person” and should be accorded human rights. Despite months of discussions with Google managers, Lemoine remained intransigent. Eventually, he was dismissed.
Another issue now widely discussed is AI’s lack of a conscience and an ability to feel guilt. Of course, the exercise of conscience is something most humans, even agnostics and atheists, share. But some AI trainers are already working on ways to train AI to simulate conscience and guilt – although, by definition, they cannot possibly experience them. This is clearly another attempt to turn machines into humanoid beings.
Another issue examined in the book is the growing problem of falsehoods and deep fakes generated by AI, often characterised by AI’s creators as “hallucinations”. This is one of the most disturbing things I examined in my research. Although these falsehoods are described as hallucinations, they can seriously mislead chatbot users.
In one prominent case, a lawyer employed by an AI company used case names in a court case that were generated by a chatbot that turned out to be hallucinations. The judge presiding over the case was not amused and sanctioned the lawyer.
Many academics have also been taken in by fictitious references supplied to them by AI. Examples of this concerned the authors of Antiqua et Nova. They point out that these hallucinations “can distort our relationship with others and with reality” and “can gradually even undermine the foundations of society”.
“As deepfakes cause people to question everything and AI-generated false content erodes trust in what they see and hear, polarisation and conflict will only grow,” the document states. “Such widespread deception is no trivial matter; it strikes at the core of humanity, dismantling the foundational trust on which societies are built.”
Fake facts can either be invented by AI or can arise from misinformation circulated by ordinary people who post falsehoods on social media. This is one of the most terrifying aspects of AI. Clearly, we all have a responsibility to avoid sharing unsupported claims and avoid taking part in social media echo chambers that only ever acknowledge one side in political debates.
Antiqua et Nova states: “Those who produce and share AI-generated content should always exercise diligence in verifying the truth of what they disseminate and, in all cases, should avoid the sharing of words and images that are degrading of human beings, that promote hatred and intolerance, that debase the goodness and intimacy of human sexuality or that exploit the weak and vulnerable. This calls for the ongoing prudence and careful discernment of all users regarding their activity online.”
The spread of AI’s influence and the future it is relentlessly forging are clearly not trivial issues. It is one of the most concerning issues facing the world today. It is not something we can put aside for another day. The most worrying thing about it is the speed with which its influence is spreading.
This is something that the people who have created the new technology – the so-called “fathers of AI” – are now also expressing serious concerns about.
They include AI luminaries such as Nobel Prize winners Dr Geoffrey Hinton and DeepMind founder Demis Hassabis. Perhaps surprisingly, they also include Elon Musk, the head of Microsoft AI, Mustafa Suleyman, and many other prominent figures in the AI world.
It is not too strong to say their visions of the future are becoming increasingly apocalyptic. Many have even signed a statement published by the Centre for AI Safety, that Artificial Intelligence could lead to “the extinction of humanity”.
Their biggest concern seems to be that the speed of change is not only likely to continue, but to accelerate. This is because businesses and governments are engaged in a race for ultimate domination of AI technology and can’t allow themselves to slow down for fear of competitors taking the lead.
Where AI is concerned, this is the most terrifying thing of all.
