EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
Hacker News·9d ago·5 min read
Things That Apparently Cause Cancer
TL;DR
Recent analyses challenge Harvard's claims linking proximity to nuclear power plants with increased cancer mortality, revealing flaws in their methodology. The findings suggest that similar proximity-based studies may yield misleading results.
✦ Why It Matters
Evaluate the methodologies of studies you rely on for health and safety decisions.
Key Takeaways
- Critically assess proximity-based health studies for validity.
- Understand the limitations of statistical models in health research.
- Be cautious of studies attributing health risks to specific locations.
- Advocate for more rigorous methodologies in public health research.
Full Summary
Harvard researchers published studies asserting that living near nuclear power plants correlates with higher cancer rates. Their methodology involved creating a 'proximity score' based on distance from these plants, treating entire ZIP codes as exposed.
Critics, including Adam Stein and Yazan Alwadi, argue that this approach leads to inflated cancer risk estimates, as it fails to account for cumulative exposure from multiple plants. For instance, their studies attributed over 120,000 cancer deaths annually to proximity to Costco warehouses and private colleges, highlighting the absurdity of the results.
The methodology's reliance on proximity without considering other factors undermines its validity, suggesting that similar studies could misrepresent health risks. These findings call into question the scientific basis for public health policies derived from such research.