The Study
AI-Generated “Slop” in Online Biomedical Science Educational Videos: Mixed Methods Study of Prevalence, Characteristics, and Hazards to Learners and Teachers
This study didn't test if AI videos make people learn worse—it just watched a bunch of videos and wrote down what looked weird or wrong. So we know what slop looks like, but we don't know if it actually hurts learning.
Analysis score
Maximum 44 for a cross-sectional study.
Where the score came from
AI can make videos that sound like science lessons but are full of mistakes, weird analogies, and boring robot voices — and people watch them just as much as real teacher videos.
Where does this study sit?
Reviews of RCTs (Meta-analyses)
Max 100Randomized Trials
Max 90Reviews of Cohort Studies
Max 85Cohort Studies
Max 72Reviews of Case-Control Studies
Max 63Case-Control Studies
Max 58Cross-Sectional & Case Series
Max 50Expert Opinion
Max 534 / 100
Quality score
Snapshots of a population at a single point in time, or descriptions of small groups. Can identify correlations and prevalence, but cannot determine cause and effect.
Key takeaways
Summary
Based on the study abstract and findings.
- 1Even though these AI videos are full of errors and confusing explanations, they're just as popular as real educational videos — meaning learners might be learning wrong things without knowing it.
- 2About 5 out of every 100 science videos on YouTube and TikTok were made by AI with little human care.
- 3These videos had the same number of views and likes as good videos.
Score breakdown, methodology, conflicts of interest, evidence analysis & raw study data
Publication
Journal
JMIR Medical Education
Year
2025
Authors
Eric M. Jones, Jane D Newman, Boyun Kim, E. Fogle
Related Content
Claims (7)
Speech generated by artificial intelligence often sounds unnatural because it leaves out words that connect ideas and creates abrupt shifts between sentences.
AI-generated biomedical videos labeled as 'slop' use inaccurate comparisons, such as likening cellular receptors to beehive gates or metabolic processes to orchestras, which distort the actual structure of biological systems and cause learners to form incorrect understandings.
AI-generated biomedical videos labeled as 'slop' include inaccurate simplifications, such as describing quaternary protein structure as interactions between multiple protein molecules and treating urea cycle disorder as one uniform disease.
AI-generated biomedical videos labeled as 'slop' use too many emotionally charged words like 'amazing' and 'crucial,' which distracts viewers and causes them to misunderstand what information is most important.
A study found that 5.3% of biomedical science videos on YouTube and TikTok contain low-quality content generated by AI with little human oversight, featuring errors such as false facts, confusing analogies, mismatched audio and video, and disorganized structure.
Videos about biomedical topics created by AI and labeled as low-quality receive the same number of daily views, likes, and comments on YouTube and TikTok as videos labeled as high-quality.
Not medical advice. For informational purposes only. Always consult a qualified healthcare professional before making health decisions.