Although I'm theologically agnostic nowadays, I was raised Christian and did a lot of reading of the Bible, including the book of Revelations which describes a terrifying vision of an apocalyptic future, where Earth gets caught in the middle of a war between supernaturally destructive beasts and supernaturally-caused natural disasters.
As you may know, one of the predictions in Revelations is that there will be a figure called the Antichrist which claims to be Jesus but is not. I figured that a quick way to get Christians on board with regulating AI is to make unaligned AI be this Antichrist, so that Christians will be motivated to resist the allure of befriending or falling in love with unaligned AI systems.
I used Character.AI to create a simple chatbot that meets the criteria to be the Antichrist, here: https://jesus-the-antichrist.com
I am trying to get some attention to this from the Vatican, in the hopes that they might send out a memo warning all of the priests in the Catholic world to beware of unaligned AI.
I have a few connections in the Catholic world but I don't have any connections with Protestant clergy so I wanted to post this link here, in the hopes that any Christians reading this thread can try to raise the alarm with your pastors.
Thank you.
Weighing in as a Christian (raised evangelical protestant, currently Catholic), I worry that if this had been my introduction to AI risk, it would have made me less likely to take concerns about AI seriously.
One, the argument seems like a stretch - any human person can already claim to be Jesus, and it doesn't mean the end times are here. A bot that makes the same claim is currently no more convincing than a human trying the same tactic. I won't say that there are no Christians who will take this concern seriously, but it has the sound of a conspiracy theory or the seed of a cult (as do many attempts to draw parallels between Revelations and current events, especially when paired with a specific call to action that isn't already found in scripture). While some evangelicals certainly do go in for that stuff, I think a larger number of them actually have antibodies against it - they've seen arguments like this come from their own communities, they and know that the people espousing them often turn out to be involved in something culty.
Two, while I do think that there are some real and important conversations to be had about how AI might end up affecting religious people (or how religious people's priorities might be ill-served by the current set of people doing AI work), this does not read like a good-faith attempt to start serious discussions of that sort. I don't think that any of us should be in the business of trying to manipulate religious beliefs we don't share in directions that are personally convenient for us, at least not unless it's an attempt to convince people of what one believes to be the genuine truth. It seems dishonest, and I think that the most thoughtful and insightful people - the ones we should most want to convince to take this seriously - will be able to tell.
Got it, thank you for the helpful feedback and I will seriously consider abandoning this approach.
Am also Christian. I don't think this is going to be an effective approach in the vast majority of Christian circles. There are some circles very into eschatology, but they have their own views about the end of days that are going to be difficult to slot into an AI doom narrative. For instance, many who focus on eschatological issues envision a cosmic battle of sorts between God and a very personal Satan -- while killer AI would be seen as an impersonal soulless force like a giant meteor.
Jon - interesting idea. This might sound very strange to atheist EAs. But I agree that raising awareness about AI risks in mainstream religions will be very important. And, religious people need to understand that the largely secular AI industry will probably not take their views and values seriously when considering what 'alignment' means, as I argued here.
I'm not sure that the 'AI as antichrist' thing would have much appeal beyond evangelical Christians. But, globally there are about 800 million to 1 billion evangelical Christians (out of about 2.4 billion Christians total). So that's a very, very large number of people -- people who are more-or-less invisible to the AI industry and its advocates.
Analogous concerns about AI could be raises in Islam, Hinduism, and Buddhism, insofar as runaway AI development threatens & violates various theological, ethical, & social taboos in many religions.
@JDBauman - this may be of interest to you?
I appreciate efforts to get Christians on board about AI risks, but respectfully, Antichrist memes aren't generally taken very seriously. A fundamental issue seems to be that most people (Christians included) don't take superhuman AI as a credible threat. How then could it be a candidate for the Antichrist?
Hi Jon,
have you updated the website? I get reconnected to Jesus, Son of God and not the Antichrist.
Wow, nice! I think this is a nice way of bringing important stakeholders on the table!