Review of Ancient Aliens S22E11: “Alien A.I.”

​I suppose it was inevitable that Ancient Aliens would eventually get around to artificial intelligence. It’s a bit surprising that it took this long, but it takes time for the lazy producers at Prometheus Entertainment to wait around until they can rip their copy from online articles without having to do any real work. The conceit of this episode is that the A.I. tools that archaeologists and historians have begun to use to analyze artifacts and data might also be used to identify evidence of an extraterrestrial presence on Earth. But most of the show is devoted to discussing recent A.I. analysis of archaeological sites without any real alien connection until the end of the episode. It’s a fine line Ancient Aliens needs to walk in this episode since it might encourage its audience to ask ChatGPT about the show and discover it is B.S., which makes it all the stranger that Ancient Aliens doesn’t bother to distinguish between ChatGPT-style large language models and machine learning—or just “computers,” for that matter.

​Segment 1
The show opens at Dartmouth in 1956 and the birth of artificial intelligence. A very quick definition and history of A.I. follows, with a slight discussion of the risks of A.I. going rogue. Giorgio Tsoukalos steps in to caution that A.I. can and does make mistakes before, in a separate clip, he talks about the “limitless” potential of A.I. to solve archaeological problems. The show asserts that A.I. is finding “long-hidden secrets,” though it declines to provide any.
 
We then move on to the use of A.I. to translate cuneiform texts, which leads the talking heads to fantasize that new Anunnaki secrets will emerge from the translations—even though they still don’t understand who the Anunnaki actually were in mythology independent of Zecharia Sitchin’s wrongheaded revisionist ideas. A.I. is starting to be used to aid in translating texts because there are not enough scholars who can read the texts—a depressing situation where our culture would rather pay through the nose to have a machine read tablets than pay a living wage for students to learn the languages needed to translate these texts. Despite Andrew Collins’s and Jason Martell’s effusive speculation, the show admits that the texts remain untranslated and no one has uncovered anything new in them.
 
All of the facts in this segment could have been stated in 30 seconds of narration rather than ten minutes of speculative, fact-free blather.
 
Segment 2
The second segment talks about how A.I. is used in searching for Maya cities. A potted history of the Maya is presented to establish how A.I. has been used to identify Preclassic cities that predate the more famous ruins of the Classic period cities. By analyzing LiDAR scans of the forests of Guatemala and Mexico, A.I. led archaeologists to tens of thousands of buildings and many unknown cities a thousand years older than the Classic period. This has opened a window into a lost phase of Maya history, but the show strains to suggest that the Preclassic period is when space aliens visited the Maya. It’s said with a shrug, however, and the majority of the segment is a straightforward discussion of current archaeological research. “A.I. is confirming much of what ancient astronaut theory is suggesting,” David Childress says, though the show is unable to provide any example of this. The closest they come is Childress’s notion that ancient astronaut theory predicted “more evidence” about the human past exists than is currently available to archaeologists. NO WAY! WHAT AN AMAZING CONFIRMATION! More evidence? Really?!?
 
Segment 3
The third segment takes us to Nazca, where A.I. analysis of LiDAR scans identified three hundred new geoglyphs. Tsoukalos mistakenly calls them “figurines,” and the show leaves this in for broadcast, presumably because no one there knows the difference between a figure and a figurine. Lots of words like “maybe” and “possibly” are bandied about to hide the fact that the new geoglyphs offer no evidence of space aliens, unless you think a whale holding a spear is a message from Zeta Reticuli.
 
We then move to Göbekli Tepe, where a potted history of the site is given, similar to descriptions provided in earlier shows that covered the same site. Andrew Collins focuses on Pillar 43, which depicts a vulture and a scorpion. Unfortunately, the show decides to adopt Martin Sweatman’s analysis of the pillar, which as regular readers know is nonsense, and the show fudges that fact that he did not use A.I. for his constellation correlations but plain old computer-aided analysis. Hugh Newman says computer-aided analysis is “A.I.-type technology,” because for this show it’s not that “everything is computer” but that “everything computer is A.I.” When Sweatman published  his most recent claim in 2024, I didn’t even bother reviewing it because it had so little evidence. Martin Sweatman alleged Pillar 43 at Göbekli Tepe was a calendar filled with the same constellations recognized by the Greeks 10,000 years later—a claim that has no supporting evidence—rooted in an assumption of precision in the carvings that is itself based on assumptions about measurements and mathematics and 2D projections of the 3D dome of the sky for which there is no proof.
 
Anyhow, Sweatman’s date of 10,500 BCE, conveniently matching Graham Hancock’s favorite date, leads Tsoukalos to claim this is proof of Zep Tepi, the term fringe history appropriated and transformed from an ancient Egyptian idea of the earliest of days.
 
Segment 4
The fourth segment looks at how A.I. is being used to reexamine the Dead Sea Scrolls, which excites Ancient Aliens because of the scrolls’ connection to the Book of Enoch, which was found among them, and its reports about the Watchers. Naturally, the Watchers-Nephilim story gets trotted out again. William Henry speculates that A.I. can somehow prove that stories of the Watchers are not Second Temple-era pseudepigraphy but rather are contemporary with the age of the patriarchs and thus are firsthand reports of meetings with space aliens. A.I. is not doing that, mind you, but they hope it might. The A.I. is trained to recognize handwriting changes over time and to use those stylistic changes as diagnostic criteria to date manuscripts. This pushes some of the scrolls back to the third century BCE—a far cry from the supposed date of Enoch, who in Biblical chronology lived before Noah’s Flood. Tsoukalos says that the redating of the scrolls demonstrates that “flying chariots” were present in the texts at the earliest phases and therefore “our ancestors” really saw flying craft, even though, of course, the Dead Sea Scrolls only refer to flying chariots as God’s and the angels’ throne-chariots in the heavens, often modeled on Ezekiel’s vision--but ancient astronaut theorists have already decided it is a UFO.
 
Segment 5
The fifth segment switches to military UFO videos and the question of whether A.I. can analyze military images better than humans. William Henry says that A.I. can process and straighten shaky videos and thus calculate what the object actually is. Travis Taylor says he hopes to create an A.I. system that, by feeding it every known UFO video, can be used to instantly analyze new videos. Garbage in, garbage out, right? He says “A.I. is a fantastic tool” that “would give you a leg up on any topic known to man,” which is a deeply depressing statement from a man once employed by the government to study UFOs, since we all know that A.I. large language models hallucinate facts and do remarkably poorly at historical interpretation. He does not differentiate between A.I. processing tools and large language models, but says that he uses A.I. every day to do his mathematical calculations for him and for “space science things” and “for the ancient astronaut theory avenue as well.” He alleges that A.I. proved to him that Teotihuacan was constructed to match the Golden Ratio. Unsurprisingly, Taylor overstates this. Similar claims for the Golden Ratio at Teotihuacan were made in the 1970s and 2000s—before A.I., using aerial photos and human brains—and upon analysis was determined to be “trivial” and “not necessarily a reflection of any deeper level of understanding or significance,” as one 2016 academic book on architecture put it.
 
Physicist Michio Kaku shows up to allege that the aliens will actually be A.I.-powered robots when they arrive here because it is “a waste of time, a waste of effort” for living beings to travel between planets.
 
Segment 6
The final segment repeats material from a previous episode about Filippo Biondi’s claim that his A.I.-powered proprietary scanning system discovered giant pillars and a lost city beneath the Giza Plateau, a claim archaeologists dismiss because Biondi has over-processed radar scans that cannot truly penetrate beneath the ground to the depth he claims nor with the detail he claims.  Andrew Collins shows up to claim that Biondi’s imaginary structures go back “forty to fifty thousand years ago,” built by “an enlightened” human population to whom we owe “all knowledge”—“the lost civilization.” There is not a lick of evidence for this claim, but when David Childress tells us that A.I. is “helping” us “find out who were are and where we came from,” you can see how the show is happy to misuse a genuine fact—that A.I. tools are used to expand archaeological knowledge—to give spurious credence to the unproven, unevidenced ancient astronaut theory’s least supportable speculations.