AI and the Christian Life with Maaike E. Harmsen

So, um, the occasion for why we address the issue of AI is, of course, that everybody's talking about it the whole time.

Um, but also more specifically the relation between Christianity and AI is the encyclical by Pope Leo, uh, that w- was published not so long ago.

It's called Magnifica Humanitas, The Magnificence of Humanity.

Um, so, and that is in, that's kind of a, I think it's roughly 80, 100-page document that gives a view from the perspective, of course, of the Roman Catholic Church, Roman Catholic social teaching on the topic of artificial intelligence.

And that, for us, was prompted us to say, "Okay, let's delve into this issue for once from the perspective of neo-Calvinism," which is, um, usually our perspective here, and that's also your field.

Uh, but maybe before we go into the details of the encyclical maybe and our neo-Calvinist reflection on it, um, maybe first a step back to ask, so what, what, what, what are we talking about exactly? What, what is AI? And I think you are very well suited to address this question because that's, I think, what you're teaching also, um, often is about.

So maybe you can just, like, take us a little bit, um, alongside in the, in taking a step back and looking at where, where are we talking about.

Yeah.

So, um, th- this, the topic of AI came about, uh, in two- 2022 when ChatGPT started, but before we had the AI, we had artificial intelligence in, in a different form.

We would call it information technology or information communication technology, and it was all, um, automated systems, uh, meant for a specific task.

Uh, depending on, you know, what you were working on, uh, it would be how Google would find your, your search history, uh, how trains would interact with each other on, uh, and how, how the, uh, um, the electric grid, um, is organized.

It's all done by automated systems.

A refinery uses software, uh, for auto- automated, uh, uh, processes.

So there, there was a lot of, um, um, information technology already around, but what was new was in 2022 was a, a generative AI, which is a actually a quite a different k- kind of, uh, uh, information technology, uh, coming abroad, which is based not on, uh, rules and commands, but based on statistics, and that's a very different b- ballgame because for stati- to make a good guess, you need, uh, a machine needs a lot of data, and to process a lot of data, you need a different kind of hardware, and there was only one company, and that was Nvidia, who, uh, would make this kind of GPUs, the, the hardware needed to, to make generative AI.

And actually when Nvidia started, um, they had a hard time finding anyone to, to buy their GPU, their, their, their, their new computer.

But, uh, OpenAI, um, and especially Musk said, "I, I would like one.

I'd li- I would like to buy them." And OpenAI started, uh, to, to, uh, process a lot of data, legal and illegal, uh, lot of copyright infringement in, in this, uh, generative AI.

And they made a, a language model that was, uh, such, so more, uh, such mo- a lot more sophisticated than we are used to.

It's not just a language model, but there are also models that can code.

Uh, Anthropic is working a lot, uh, with that, um, but also can find, um, cybersecurity threats.

There is, uh, uh, there are models that can make video.

There are models that can make worlds.

Uh, so you can, you can use it for, um, making cars safer.

Um, and all this we call now AI.

It's actually the second form of AI from our normal information technology, which we, we, which we already had around since, since the 1960s.