The AI Priest

Catholic Answers, an organization that runs a website dedicated to promoting Catholicism, created and released an AI model last week that answers questions about Catholicism. The AI model, originally named Father Justin, provides a human-like interaction through a desktop browser in which people visiting the AI can ask it questions and receive human-like responses.

The catch is that the AI model answered questions as if it was REALLY a priest, according to the author of this article at Futurism. That is, the AI talked as if it was a real priest, even stating that it lived in Italy and that it was drawn to the priesthood (i.e., becoming a priest). Of course, these are things that an actual priest might really say, so the AI is, in that sense, actually mimicking a real priest fairly well. But the fact that it is not a real priest was a cause for concern for some Catholics and for Catholic Answers, so much so that they adjusted the model so that it was now longer portraying itself as a priest. It is now just "Justin" instead of "Father Justin."

Read the article at Futurism and think about the potentials and pitfalls of using AI to promote a religious group. In particular, ask yourself:

• In what ways can AI be used to help a religious group carry out its mission?
• Would some religious groups be more open to using AI than others? Which ones would be more open to using AI, and why?
• What kinds of individuals would be most likely to benefit from an AI like Justin?
• What are some of the dangers for a religious group from using an AI like Justin?

COVID-19, Religion, and Survey Responses

Many studies have provided evidence that religious participation declined during the COVID-19 pandemic. However, a recent article published in Sociology of Religion (one of the leading journals that publishes social-scientific research on religion) shows that the purported decline may be due to a change in survey collection and not a change in religiosity.

The article is too long to read in its entirety, but you can skim it at the link above (use a computer from on campus or use the UCI VPN to access it). 

The gist of the paper is as follows. The General Social Survey (GSS), which is the most frequently used survey in religion research, changed its collection method during the pandemic from face-to-face interviews to online. This change in mode of data collection resulted unintentionally in a change in the types of people who agreed to participate in the survey. In particular, highly religious persons were less likely to participate in the new online format, and this resulted in a decline in the measured religious participation in the data from before the pandemic.

This finding teaches us an important specific lesson about research on religion and how we must always be cautious about the sources of data that we use. Researchers must strive to be aware of changes in the method of data collection when comparing different datasets.

It also tells us something important about American religion, namely, that American religion did not decline in the way that many people thought it did during the COVID-19 pandemic. In hindsight, it might not be surprising that religion did not decline as some researchers had concluded because people often turn to religion during difficult times. This paper provides the evidence to support the hunch.

Finally, it provides an important cautionary tale to future researchers who want to design religion surveys. Every survey mode has pros and cons, and it is important to be aware of those pros and cons when designing your survey.

Religion, Free Speech, and Hate Speech

When does saying something negative about a religious group constitute hate speech?

Well, you first have to define hate speech, and this is not as easy as you might think. The basic idea is that hate speech has two parts:

1. It is language that targets an individual or group based on an inherent characteristic (religion, gender, race, etc.).
2. It has some sort of aggression or even potential for harm or violence.

Part 1 seems obvious, but 2 is more difficult to pin down because it is not always clear if and when statements are harmful. Should saying something negative about a religious group constitute hate speech just because it is negative? Can't any sort of negative speech have the potential to promote aggression or violence even unintentionally? Note that there is actually no legal definition of hate speech under U.S. law, however it can be a criminal act to use speech or other expressions to incite criminal activity or threaten violence against a person or group.

This background helps us to contextualize the findings from a recent Pew Research Center survey about religious discrimination in the U.S. Americans report that there has been an increase in discrimination against both Muslims and Jews since the start of the Israel-Hamas war.

Check out the page, and especially examine the charts. There is a lot there, but you can focus on just the first findings reported which relate to the possibility of hate speech. The connection to hate speech is that there have been accusations of hate speech against supporters of both sides of the war.

Notice that there is not a perfect consensus among Americans about what types of speech should be allowed, but of course a consensus is hard to expect. There are, on the other, opinions that a majority of Americans hold.

First, 73% of American's say that speech that calls for violence against Muslims or against Jews should not be allowed. That is, a large majority of Americans believe that calls for violence should not be protected as free speech.

Second, there is a lower but still majority amount of support (about 60% give or take) for allowing speech that opposes Israel's right to exist as a Jewish state or that opposes a Palestinian state. This type of speech is more difficult to assess for Americans because it is negative and could be viewed as implying support for violence against Israel or Palestinians but does not necessarily have to be interpreted as implying support for violence.

Nonetheless, the survey results reveal that most Americans see a big difference between saying something that could be construed as potentially conflictual in nature (most Americans believe this should be allowed) and saying something that is explicitly violent (even more Americans believe this shouldn't be allowed).

This distinction can also help us understand the reasoning and intent behind religious hate crime laws that are meant to protect religious minorities. As we will discuss later in the quarter, the enactment of these laws can help to protect members of religious minorities from violence and therefore help to maintain or increase the religious diversity within a country.

Religion and Happiness

Empirical results from social-scientific data analysis are always under scrutiny, but there is one pattern that is so well established that it appears to be beyond dispute. That pattern is the strong and robust positive relationship between religion and happiness. A recent article in the Deseret News provides a very accessible summary of some of the findings.

Read this article, and as you read you should ask yourself a variety of questions.

• Why are religious people happier?
• Is there a causal relationship between religion and happiness so that religion actually makes people happier?
• Are all religious people happy? Why or why not?
• If religion does make people happier, then does mean that religion is overall good for society?

A Continuing Trend in Government Restrictions on Religion

Last month, the Pew Research Center released a report on changes in government restrictions on religion around the world. Read the summary here. If you are interested, then you can also check out the full report pdf, but be warned that it is pretty long.

The key finding is that the median level of government restrictions on religion increased slightly from 2020 to 2021. This development is a continuation of a trend that the Pew Research Center has been measuring since 2007 using their Government Restrictions Index (GRI).

In statistics, an index is a composite statistic. That is, it is a single number that summarizes information from several data sources.. Creating an index is no simple thing because it can be conceptually difficult to determine how best to combine different data into a single number, but indices are very important in the social sciences because of their ability to summarize many things at once. And this is true for the GRI. Because government restrictions can take many forms, the GRI provides a single number that summarizes all of those government restrictions in a useful way.

As you read this summary, you should look for information on the GRI, what it measures, and how it measures it. Also look for information about how government restrictions are related to other types of government interventions, such as the government's providing of benefits to religious groups.

Once you've read the summary page, then click on the link that takes you to Chapter 4 to learn about restrictions in the world's 25 most populous countries. Scroll down to see the graphic. Which large countries have the most restrictions? Which large countries have the fewest restrictions? Any surprises?