"AI sycophancy carries genuine risks when LLMs are used in settings such as health care ... McCoy and his team reported that LLMs used for medical reasoning often changed their diagnosis when physicians added new information, even if the new inputs were irrelevant to the condition."

"LLMs have been trained to overly agree with humans or overly align with human preference, without honestly conveying what they know and what they do not know,” says Gao. What is needed, she adds, is for the tools to be retrained to be transparent about uncertainty.

However, in many Christian plurality countries, we have seen a lot of unaffiliated growth via switching in recent years.

Well, the impact of switch vs. fertility is complicated. China has by far the most unaffiliated people. Since about 90% of the country is unaffiliated, there’s limited potential for growth via switching. However, low fertility is expected to significantly shrink China’s population.

We've done a little on this, and it would be great to do more. Here's a 2022 report on knowledge about international news (obviously many of these items are outdated now): www.pewresearch.org/global/2022/...