TRT and blood pressure: what the evidence actually says

What did @ogslimyfella actually say?

The creator found a ChatGPT prompt circulating in the comments of a "black pill" edit on TikTok. The prompt is designed to give users an AI-generated attractiveness rating based on a photo. After using it, he was apparently satisfied enough with the result to declare "y'all can't tell me shit" and promise to share the prompt with followers. His confidence, he made clear, comes from an AI's assessment of his appearance.

To be fair, this is not a health claim. It is not TRT advice. It is a man who got a good score on a novelty AI test and wanted to share the moment. That context matters before we start picking it apart.

Does the science back this up?

No meaningful science backs the idea that AI chatbot prompts can reliably rate human attractiveness. Full stop. Large language models like ChatGPT are not trained to evaluate facial aesthetics with clinical or even consistent accuracy. They generate plausible-sounding responses based on pattern matching, not perceptual judgment.

Researchers have documented that AI systems evaluating faces tend to encode significant demographic and racial bias. A 2019 study by Buolamwini and Gebru (MIT Media Lab, published in the Proceedings of Machine Learning Research) showed that facial analysis systems had substantially higher error rates for darker-skinned women. Attractiveness-rating tools carry the same structural problems. Any score a chatbot gives you reflects the biases baked into its training data, not an objective read on your face. There is no peer-reviewed standard for AI attractiveness scoring. The number it gives you is, in technical terms, made up.

What did they get wrong (or right)?

He got the entertainment value right. This is a fun, harmless social media moment and he is not pretending otherwise. He even says "I just hope it doesn't ruin your day," which is a more responsible framing than most viral AI rating content.

Where things get worth flagging is the broader ecosystem this content feeds into. The "black pill" community, which he references directly, promotes deterministic and often toxic beliefs about male attractiveness, social dominance, and genetic hierarchy. AI rating prompts are frequently used within that community to either confirm or devastate self-worth based on a number generated by a language model that cannot actually see you the way a human brain processes a face.

Psychologically, this is not neutral. A 2022 study by Vogel et al. in the Journal of Social and Clinical Psychology linked appearance-focused social media engagement to increased body dysmorphic symptoms in young men. Handing that anxiety loop over to an AI does not make it less damaging.

What should you actually know?

If you are here because this video showed up on your TikTok and you are curious about testosterone, hormone optimization, or any TRT-related content, here is the actual relevant information. Physical appearance, including body composition, skin quality, and energy presentation, can be affected by testosterone levels. Low testosterone is associated with increased body fat, reduced muscle mass, and fatigue, all of which affect how someone looks and feels.

But an AI chatbot cannot assess any of that from a selfie. It cannot detect your testosterone levels. It cannot evaluate inflammation, metabolic health, or any biomarker that actually predicts physical vitality. A blood panel can. If you are a man in your 20s or 30s noticing changes in body composition, energy, or mood, the useful step is lab work, not a ChatGPT prompt. Total testosterone, free testosterone, LH, FSH, and SHBG give you real data. An AI "rating" gives you a number that says more about its training data than about you.