Nowhere to Stand

We like to think of ourselves as independent thinkers. In the information age, the universal battle cry across the political and social spectrum is to "do your own research." But this ideal has become a structural impossibility.

One of the defining characteristics of modernity is that you are forced to trust institutions. None of us can be the expert on the myriad complex systems that we touch in our day to day from drinking-water filtration to sleep-tracking wearables; you cannot smell a carbon emission, you cannot see a spike in a pathogen's Rt value. The normal experiential logic of daily life is such that our immediate reality is dictated to us from the top down by chemical assays, medical consensus, and twenty AI researchers at a frontier lab in the Bay.

When we attempt to reclaim autonomy by taking an Instagram or YouTube crash course in what is otherwise a highly specialized field, we mistake access to information for genuine understanding. Because if experts in the field can disagree, despite having specialized tools and access to highly esoteric knowledge, then how can we as laypeople form a meaningful point of view about which of the experts is reading the material incorrectly? C. Thi Nguyen has this helpful notion of cognitive islands: domains where judging who even counts as an expert requires the very expertise you're trying to borrow, so the newcomer has nowhere to stand.

People nevertheless try. For example, we like to cosplay as meta-experts and resort to techniques that don't require us to dabble in the actual subject matter: by asking who has skin in the game, who has conflicts of interest, who was most compelling stylistically in their explanation?

But those are the wrong instincts. Compelling explanations are precisely what a doubt campaign is engineered to deliver — and the rest is gameable too: an astroturfed "independent expert" is built to pass your skin-in-the-game and conflict-of-interest checks. Oreskes and Conway traced this exact behavior in Merchants of Doubt: the tobacco and leaded-gasoline campaigns ran on planted experts and captured journals, plus the all-purpose cover of "just asking questions."

So if you are a non-expert, a highly sophisticated false argument will sound just as convincing, elegant, and airtight as a true one. If a brilliant contrarian presents a granular theory on why peptides are going to make you gain lean muscle mass 40% faster, you lack the specialized tools required to find the hidden flaw. You will nod along, find it completely plausible, and be utterly defenseless against being conned.

So are we just left with Scott Alexander's epistemic learned helplessness? Not quite. Because even that leaves open the question of which dogma and which institutions to anchor your faith in. There have been some proposed escape hatches here: For instance, you might ask, does this institution frequently correct its own errors and march towards progress or does it punish its own dissenters? But, that too is a question a layperson cannot actually decipher. Because even a captured field can have a "peer-review" process, retraction machinery that functions as pageantry, and all the other bells and whistles. Auditing whether self-correction is real or performed is itself expert-level work. No institution is clean enough to outsource your judgment outright. You stay exposed.

"Do your own research" was always the wrong instruction. The mature response is neither confidence nor surrender, but something harder: placing a provisional, revocable trust while admitting you can't fully justify where you placed it. You need to hold your certainties loosely enough that the next correction doesn't break you.