Mass Exposure Paradox
post by max-sixty · 2025-04-16T20:18:00.492Z · LW · GW · 0 commentsContents
Examples Mass Exposure as Meta-Evidence Memetic Fitness Fueled by Universal Relevance Similar Concepts Spend Your Energy on Other Things None No comments
Why are so many health-related anxieties related to exposures that we all encounter?
Here's a model that offers some explanation — I'm calling it the mass exposure paradox. It arises from two opposing consequences of universal exposure:
-
Mass exposure to something harmful would generate highly significant evidence for its harm. When billions of people use or consume something for decades without a clear epidemic of harm, it becomes strong evidence against large effect sizes. Genuinely harmful exposures — such as leaded gasoline or cigarette smoking — leave unmistakable signals.
-
Mass exposure boosts memetic fitness. When everyone is exposed, narratives of harm can gain cultural traction, sweeping through social networks and broadcast media.
Examples
- Cellphones & brain cancer: pervasive use since the 1990s; no solid tumor spike, but fears linger.
- Aspartame: in foods since the 1980s, extensively studied; fears persist despite reassuring evidence.
- Vaccines: centuries of use; exceptional safety record despite rare adverse events; we still have large anti-vax movements
- Other examples: GMOs, WiFi/5G, seed oils, "toxins", microplastics.
In each case, mass exposure provided a natural experiment demonstrating safety, while its memetic fitness allows anxiety to spread and persist.
Mass Exposure as Meta-Evidence
The very ubiquity of an exposure strongly implies it's safe. If billions of people have been exposed for decades yet we still debate whether any real effect exists, the effect size must be negligible or nonexistent.
Genuinely hazardous substances which had widespread exposure (lead, asbestos, smoking) show clear, undeniable population-level harm without cherry-picking studies. We can generally rely on intensive research scrutiny finding the real dangers of widespread exposures if they exist.
In contrast, if you're worried about harm from exposures, it's much more likely to come from something you're specifically exposed to — something in your town, related to your behavior, or affecting people with your genes — because we don't have nearly as much evidence on those things.
Memetic Fitness Fueled by Universal Relevance
Why do these low-evidence fears flourish and persist? Precisely because exposure is universal:
- Shared context: Everyone is exposed, making anxieties easy to share. Economies of scale in media mean that it's worthwhile to, say, do a podcast on the dangers of vaccines, because there's a large audience who are considering vaccines.
- Identity markers: Avoidance behaviors (EMF shielding, anti-fluoride stances) reinforce group identity. They allow people to signal their conscientiousness and their locus in a social network.
- Science aesthetic: There's an attractive aesthetic of scientific inquiry (this is quite different from, say, astrology) that allows signaling of competence and intelligence.
- Pre-existing distrust: Narratives from the left about corporate greed, or from the right about government conspiracies, easily resonate.
- Resistance to falsification: Claims with diffuse causality like "effects take decades to manifest" or tenuous connections like "vaccines cause autism" resist easy debunking.
So, even faint signals sustain fear precisely because everyone is involved.
Similar Concepts
There are plenty of theories on the spread of low-information issues, such as availability cascades or moral panics.
The novel thing here is that universal exposure both undercuts real risk and simultaneously fuels fear (though interested to read anything that has highlighted this previously).
Spend Your Energy on Other Things
With that conceptualization, we should recognize that there will be many false claims of this category and so our prior on something in this category being true should be very low.
And, unless you're a specialist in the specific area or you happen to be specifically interested, I would claim that you should dismiss most reports with these properties as suffering from this paradox.
0 comments
Comments sorted by top scores.