AI Biomechanical Analysis of Patterson-Gimlin Film Reveals Anomalous Stride Ratio

So there's a video making the rounds on YouTube that dives into something pretty fascinating — an AI-powered analysis of the Patterson-Gimlin film that apparently produced results nobody was quite ready for. The folks over at the Reef Discovery channel put together a deep dive into how Elon Musk's Grok AI was tasked with examining the clearest scan of the original 1967 footage ever produced, and what came back has researchers buzzing. Here's the backstory for anyone who needs a refresher. On October 20th, 1967, Roger Patterson and Bob Gimlin were riding horses along Bluff Creek in Northern California's Six Rivers National Forest when they came around a bend and encountered something that changed everything. Patterson had a 16mm Kodak camera in his saddlebag — he had been actively searching for evidence of Sasquatch for years — and he dismounted, grabbed the camera, and ran toward the subject while filming. The resulting footage, roughly a minute and 15 seconds long, shows a large, dark-haired, upright figure crossing a gravel bar along the creek, pausing to glance back over its right shoulder at what's now known as Frame 352, and then disappearing into the tree line. That moment has been debated, dissected, debunked, and defended for nearly six decades. Disney animatronics engineers in the 1970s concluded the suit technology required to produce the movement visible in the film simply didn't exist at the time. Soviet scientists — yes, during the Cold War — analyzed the footage and concluded the figure demonstrated primate locomotion characteristics inconsistent with a human operator. Primatologist John Napier, one of the most respected physical anthropologists of the 20th century, examined the film and stated publicly that while he couldn't confirm the subject's identity, he couldn't dismiss it either. Coming from a scientist of Napier's stature and natural caution, that was a significant statement. The debunkers have had their shot too. Bob Heironimus claimed to be the man inside the suit, but his account contained enough factual inconsistencies with the documented record that it raised more questions than it resolved. The suit he claimed to have worn has never been produced. Philip Morris, who he claimed made the costume, initially appeared to support the claim before his account evolved in ways that undermined its credibility. The confession never closed the file — and if it had, nobody would still be talking about this footage 57 years later. What makes this particular video worth watching is the methodology behind the AI analysis. A research group commissioned Grok to perform a biomechanical analysis using photogrammetric scans of the original film frames — not copies, not transfers, but reconstructions from the original photographic emulsion. Photogrammetry, the science of extracting precise three-dimensional measurements from two-dimensional images, has advanced dramatically in the last decade. Applied to high-resolution scans of the Patterson-Gimlin frames, it allows for reconstruction of three-dimensional spatial relationships between body parts across the movement sequence with a precision that simply wasn't available to any previous analyst. The kicker? They instructed Grok explicitly to perform its analysis without a target conclusion. No preferred outcome. Just the data, cross-referenced against extensive databases — thousands of documented gait analyses across human body types, movement analysis from chimpanzees, gorillas, and orangutans, and documented movement from full-body suit productions with far larger budgets and more sophisticated fabrication capabilities than anything available in 1967. The output came back in a format that the research group spent several days working through before releasing any summary. Not because it was ambiguous — because it wasn't. The frame-by-frame biomechanical analysis produced a series of measurements that individually might each be explained within normal variation, but together form a profile the AI assessed as consistent with one category of subject and inconsistent with two others. The video gets into specifics about double stride length ratios, muscle movement visible beneath the body covering, the proportional relationship between limb length and torso, and the gluteal definition visible in certain frames — problems that Hollywood costume designers have consistently said they cannot identify a mechanism to produce, even today. As one practical effects artist reportedly put it: you cannot fake muscle you do not have. This is the kind of analysis that pushes the conversation forward. For anyone who's spent any time in this community, the Patterson-Gimlin film isn't just a piece of footage — it's the cornerstone of the entire field. Every new tool, every new methodology, every fresh set of eyes brings something to the table. And an AI without professional investment in any particular conclusion, working from the cleanest data ever available, is exactly the kind of outside perspective this question has needed. The video is worth the watch. It walks through the history of expert examination, the limitations of human analysis applied to ambiguous evidence, and what happens when you remove the preconceptions and let the data speak. Check it out and see what you think.