Grok AI Analyzes Patterson-Gimlin Film, Finds No Costume Evidence
Posted Tuesday, August 18, 2026
By Squatchable.com staff
# Grok AI Just Put the Patterson-Gimlin Film Under the Microscope and the Results Are Raising Eyebrows
So this video popped up recently, and honestly, it's one of those pieces of content that makes you sit back and think for a minute. Someone fed the legendary Patterson-Gimlin footage into Grok, Elon Musk's AI system, and asked it to do something no human analyst has been able to do definitively in nearly 60 years: figure out what the heck Patty actually is.
The results? Wild doesn't even begin to cover it.
For those who need a refresher, the Patterson-Gimlin film was shot on October 20, 1967, at Bluff Creek in Northern California. Roger Patterson and Bob Gimlin were riding along a dry creek bed when they spotted a large, bipedal, hair-covered figure near some fallen logs. Patterson grabbed his 16mm camera and captured roughly 59 seconds of footage as the subject walked away and glanced back once before disappearing into the timber. The figure became known as "Patty," and that famous turnaround moment is what researchers call Frame 352.
Now, Grok didn't just give a yes or no answer. According to the video, the AI ran frame-by-frame biomechanical analysis, motion tracking, texture correlation, lighting modeling, and comparative anatomy on high-resolution digital scans of the original film. It measured limb proportions, calculated the center of gravity, examined joint articulation, and specifically scrutinized Frame 352 for any signs of mask edges, seams, or artificial boundaries.
Here's where things get interesting.
Grok reportedly found no evidence of digital manipulation. Whatever was on that film was physically present when Patterson hit record. The biomechanics didn't match a human in a suit either. The gait, arm swing, hip rotation, shoulder movement, and balance characteristics showed none of the compensatory patterns humans display when trying to mimic an unnatural walk. No awkward corrections, no visible tension, no instability.
The center of gravity came back higher and more forward-shifted than a human's, which would be consistent with a heavier, more massive torso. The arm-to-leg ratio fell outside normal human parameters. The mass distribution suggested a structure that would have been incredibly difficult to fake with 1967 costume technology.
And then there's the fur analysis. Grok detected movement patterns consistent with real hair shifting over skin, with subtle surface displacements that matched underlying muscle activity. In Frame 352, no mask edges, no seams, no artificial transitions between face and fur. The face appeared continuous with the body.
So what's the verdict? Grok couldn't identify the subject. It didn't match a human, didn't match a gorilla, didn't match a bear, and didn't fully align with any known species. The AI concluded with high confidence that the Patterson-Gimlin film captured a large bipedal biological entity that doesn't fit recognized taxonomy. It couldn't call it Bigfoot because Bigfoot isn't a scientifically recognized species, but it stated the subject appeared real, physical, and inconsistent with a costume or special effects.
That's a pretty significant statement coming from an AI system designed to analyze data objectively.
Now, skeptics will absolutely push back on this, and that's fair. AI can be fooled, and a verdict of "unknown" isn't the same as discovering a new species. But this is arguably the most sophisticated analysis ever applied to that footage, and it makes the costume theory a whole lot harder to defend.
The video breaks down the full history too, covering the Native American accounts of Sasquatch from Pacific Northwest tribes like the Salish, the settler and lumberjack reports from the 1800s, the 1958 Bluff Creek trackway discovery that put Bigfoot on the map, Patterson's obsession with proving the creature real, and all the back-and-forth between believers and skeptics over the decades. It even touches on the Greg Long book and Bob Hieronimus claims from 2002 that tried to expose the film as a hoax, and the problems that emerged with those stories.
Honestly, this one is worth the watch. The way the AI findings are presented, combined with the historical context, makes for a compelling case that the Patterson-Gimlin film is even more mysterious than people realized. Whether you're a longtime researcher or just curious about the subject, this video brings something new to a decades-old debate.
Check it out and see what you think. The mystery of those 59 seconds at Bluff Creek isn't going anywhere, but Grok just made it a lot harder to dismiss.