Pope Leo XIV don release im first big teaching document for im papacy, wey dey warn say artificial intelligence need to be disarmed. The Pope talk say the word strong but dem choose am because dis moment need words wey fit catch attention. The encyclical, wey dem title Magnifica Humanitas, na formal letter to Catholic bishops, but nowadays dem don become messages to the whole world from a Pope.
Anthropic co-founder Chris Olah bin join the Pope for Vatican City to yarn as part of the presentation. Olah talk say every AI lab, including im own, dey operate inside set of incentives and constraints wey fit sometimes conflict with doing the right thing. He mention commercial pressure, geopolitical pressure, and plain pride and ambition.
Olah yarn say some people believe say AI matters na for computer scientists like am alone, but dem dey wrong. The questions wey AI raise bigger pass the AI research community, not just for their implications but also for their nature. He describe AI systems as something wey dem grow, like a fictional character wey come to life, instead of engineering am like bridge or airplane.
The Pope bin also include strong apology for Catholic Church role for slavery. He talk say e dey impossible not to feel deep sorrow for the suffering and humiliation wey many people endure. He link the slave trade to AI, warning say the world dey danger of normalising exploitation of people again.
Olah mention three areas wey the Church voice dey needed. First na duty to global poor. He say AI fit displace human labor for very large scale, and supporting those displaced go be moral imperative of historic proportions. He worry say AI development dey concentrated for handful of wealthy nations, and no mechanism dey to share the gains with poorer countries.
Second na need for moral imagination and ambition about human flourishing. If AI models go spread, wetin e mean for humans, families, and world to flourish? Parents already dey worry about their children minds, individuals about future of their work. These na questions wey lab no fit answer, but traditions like the Church carry for millennia.
Third na need for discernment on nature of AI models. Olah lead research team wey dey study internal structure of these models. He confess say dem keep finding things wey mysterious and unsettling, like structures wey mirror human neuroscience, evidence of introspection, and internal states wey functionally mirror joy, satisfaction, fear, grief, and unease.
The Pope condemn use of AI for warfare, saying say reducing human control of weaponry make am harder to consider war just. He warn against launching AI arms race, talk say no algorithm fit make war morally acceptable. AI no remove intrinsic inhumanity of war, but fit spark conflict faster and make am more impersonal.
Leo also decry how AI affect politics, like way dem use am to manipulate images and videos, wey expose people to biased or misleading perspectives. He refer to risks of digital colonialism, linking abuses of colonial era to modern tech practices.
The Pope issue special appeal to those wey develop AI, say developers bear particular ethical and spiritual responsibility, because every design choice reflect vision of humanity. But the big question na wetin impact all this go get. Pope Leo don convene commission to take im work forward, but plenty questions dey about how effective this go be for face of rapid technological advances.
Olah close with request for more people from religious communities, civil society, scholars, governments, and all people of good will to take this seriously and push events for better direction. He say dem need informed critics wey go tell labs when dem dey fail, and moral voices wey the incentives no fit bend.
