Recently I had some time to spare and wondered how artificial intelligence—now revolutionizing everyone’s information landscape—deals with Jesus mythicism. I use Copilot, owned by Microsoft, and began to type in the question: “What is the general opinion today regarding Jesus mythicism?” But I can pretty much guess the answer to that question, so I decided to get specific and instead asked Copilot about the uproar that took place almost twenty years ago, after BAIAS (Bulletin of the Anglo-Israel Archaeological Society) published my paper critical of then-current fieldwork in Nazareth. (TOC and entire 2008 volume available for download HERE.) My paper elicited several published reactions in the same issue. What follows is the Nazareth-related material in the BAIAS 2008 issue:

• René Salm, “A Response to ‘Surveys and Excavations at the Nazareth Village Farm (1997-2002): Final Report’ ” (pp. 95–103).

The responses to my above article (in the same issue) were:

• “On the Nazareth Village Farm Report: A Reply to Salm” (S. Pfann & Y. Rapuano, pp. 105–08);
• “Nazareth Village Farm: A Reply to Salm” (Ken Dark, pp. 109–11); and
• “The Nazareth Village Farm Project Pottery (1997 -2002): Amendment” (Y. Rapuano, pp. 113–35).
For good measure, the BAIAS issue also included Ken Dark’s hostile review of my recently-appearing book, The Myth of Nazareth: The Invented town of Jesus (pp. 140–45).

Well, it appears that history has “alternative facts.” I typed into Copilot, assuming the third person: “Was Salm published in Baias?” Here is what the AI platform replied:

Huh? What happened to my article? Now I know what it feels like to be obliterated from history…

So I asked Copilot: “Something’s not right. I read somewhere about an article by Salm (I think it’s mirrored on Academia.edu) in BAIAS 2008, something about the “NVF”.” Copilot’s response:

Wow! Ken Dark wrote my article about the Nazareth Village Farm?! 🙂 To make sure, I asked Copilot:

Gee… Ken Dark wrote the “full archaeological article, not a book review.” So one of the premiere AI platforms is telling the world that Dark did write my article! 😩 Yet the full BAIAS issue is online (above link) for anyone to see. Did Copilot not check an obvious source?

Now I was getting angry… A little further on in the conversation, Copilot wrote:

Every one of the above points is false. Especially creative is the part that people “often quote or reproduce tables/figures from Dark’s BAIAS article.” There are no tables/figures in Dark’s three page article—but there are in Rapuano’s “Amendment.” And about Academia.edu, to date I have not uploaded anything on the NVF to that platform. (Now I just might.) Copilot seems to be getting everything wrong. I then asked the AI the following:

Yowza. Copilot thinks that Rapuano critiqued Dark! But Salm critiqued Rapuano, and Dark critiqued Salm… How bungled up can AI get? Copilot’s sequence of publications that followed the above snippet:
1) Ken Dark publishes an article on the Nazareth Village Farm (NVF)
2) Dark also publishes a review of Salm’s book
3) (In BAIAS 2009, Vol. 27) Michael Rapuano publishes a critique of Dark’s NVF article.

The first point is half-right: Dark didn’t publish directly on the NVF but on my input: “Nazareth Village Farm: A Reply to Salm” (BAIAS 2008, pp. 109–11). And, yes, Dark did publish a review of my book (point 2). But the third point is very strange. Rapuano is not mentioned at all on the BAIAS 2009 content page. And if he wrote such an article (which he didn’t), it certainly wouldn’t be “a critique of Dark’s NVF article”! (Incidentally, Rapuano’s first name is Yehudah, not Michael.)

If it’s possible, things now get even weirder… I have not been able to make heads or tails out of the following snippet that ensued in this Copilot conversation:

I used Rapuano’s arguments against Dark??? Rapuano was all wrong on his pottery analysis—after all, pursuant to my article above, he had to write an “Amendment” to his NVF pottery report. I wouldn’t use anything Rapuano wrote against anybody. And, to set the record straight:
– No, Dark did not write the NVF article in BAIAS.
– No, Rapuano did not critique Dark in BAIAS.
– And no, Salm (online) did not critique Dark using Rapuano’s arguments.

To make a long story short(er), I finally presented Copilot with the link to the table of contents to the 2008 issue of BAIAS. Copilot’s reaction to this astonishing information follows:

I wasn’t satisfied and wanted to understand why Copilot didn’t check the BAIAS issue, which is right there online. Here is my question and Copilot’s (lame) answer:

The root of the problem is Copilot’s admission: “I relied on a widely repeated secondary narrative instead of checking the primary source.” Ughhh!

The question now becomes: “Why?” Why does AI not check “the primary source” and instead relies on “a widely repeated secondary narrative” that is oh, so wrong? The only reasonable answer I can give is built-in bias. In some cases, the algorithms that AI use favor false narratives over actual fact. That is very onerous in the case of Jesus mythicism. It looks like AI will not be a friend to Jesus mythicists.

Equally sad is that Copilot does not learn from user input:

So, you can “correct” AI all you want, but it won’t remember… As soon as you close the page, AI is back to where it was at the beginning of your conversation.

So there you have it: Beware of AI!

PS–I welcome comments/observations from readers regarding interactions with AI and mythicism. Do send them along! 🙂