Nadia Asparouhova on antiemetics: why good ideas resist spreading and how obscurantism is information warfare

Summary

Nadia Asparouhova's new book, Antiemetics, starts from a simple but underexplored premise: not all good ideas spread. The conventional assumption is that quality is sufficient — that a compelling idea will find its audience. Asparouhova argues there's an entire class of ideas, anti-memes, that resist sharing and remembering even when people find them genuinely interesting. Taboos, cognitive blind spots, and socially awkward truths all qualify.

The concept originates in horror sci-fi. Asparouhova traces it to There Is No Antimemetics Division by Qntm, published in 2021, which imagines anti-memes as anthropomorphized creatures that consume everything they touch but leave no memory of the encounter. She read it during Covid, was working at Substack at the time, and spent the years since trying to apply the concept to the real world.

Public channels vs. private chats

One of the book's central arguments is that the retreat from public social media into group chats hasn't solved the information problem — it's mutated it. Group chats aren't safe havens; they're incubators. Ideas get workshopped by small, aligned groups and can emerge wilder than they went in. Asparouhova's framing is that public channels and private channels feed each other rather than replace each other, with ideas cycling between the two on a timeframe that appears to be compressing.

Obscurantism as information warfare

The most practically pointed section of the conversation concerns whether anti-mimetic properties can be manufactured deliberately. Asparouhova says yes, and the mechanism is counterintuitive: making an idea boring is more effective than locking it down. Hard suppression creates curiosity; making something tedious or difficult to parse causes people to wander away. She frames this as obscurantism, citing Nick Bostrom as the source of the term, and describes it explicitly as an anti-mimetic information warfare tactic. The observation about X dynamics fits the same logic — a long, dense response to an attack tends to end the conversation by exhaustion rather than persuasion.

Anti-mimetic vs. anti-commercial

Asparouhova draws a distinction worth preserving. Some ideas fail to spread not because of any intrinsic property but because no commercial engine exists to push them. Intermittent fasting is the example on the table: if the answer to a chronic health problem is "eat less," there's no product to sell and therefore no marketing budget behind the idea. She argues this is a separate failure mode from genuine anti-memetics, though the practical outcome — suppressed diffusion — looks similar. Health is the domain where she sees this most clearly, with chronic conditions that affect large numbers of people but whose answers can't be monetized.

AI as an anti-meme surface

On whether a well-trained LLM could surface ideas that humans systematically suppress, Asparouhova is cautiously optimistic at the individual level. People already use AI to extract honest self-assessments that friends won't give. At the network level she's more skeptical — getting a taboo idea to spread from person to person remains hard regardless of whether one individual fully believes it.