I couldn't agree more. This is the part the tech-bro biohacking crowd completely glosses over. Carl Jung wasn't just being poetic; he was clinically observing patients whose psyches were fracturing under the weight of uncontained archetypal material. In Psychology and Alchemy, he writes about the danger of "inflation" where the ego gets fused with a powerful archetype, but the flip side is exactly what you're describing—the ego getting dismantled before it's ready. It's not enlightenment; it's disintegration.

My own crash course in this came during a brutal breakup a few years back. I turned to tarot like it was a diagnostic tool, but I was emotionally raw. I kept pulling cards that felt like sledgehammers to the ribs. I remember pulling The Tower and Death back-to-back while I was already dissociating. As I wrote in a previous post, "Sometimes it just turns on the floodlights in a room you weren't prepared to enter." I didn't have the grounding to integrate that. Now, I treat the cards strictly as journaling prompts. I ask, "What's the shadow content here?" rather than "What's going to happen?" It's safer. It keeps the ego intact enough to do the work without getting swallowed.

You're spot on about the survivorship bias in the literature, too. The 2012 study on cortical thickness is frequently cited as definitive proof of meditation's neuroplastic benefits, but it relies on a remarkably small sample size and lacks a matched active control group. As subsequent meta-analyses have pointed out, early neuroimaging papers on contemplative practices often overgeneralize findings from highly experienced practitioners to novices. When you combine those methodological limitations with the psychological risks of unguided introspection, the clinical picture becomes far more nuanced. We're seeing a growing body of literature—such as Van Dam et al.'s (2018) systematic review on adverse effects of intensive meditation retreats—documenting how prolonged, unmonitored practice can trigger anxiety, depersonalization, and even transient psychotic episodes in susceptible individuals. The takeaway isn't to abandon these practices, but to contextualize them. Structured protocols, qualified supervision, and a clear assessment of baseline psychological resilience are essential. Meditation isn't a neutral cognitive exercise; it's an active modulation of attention and affect that warrants the same methodological rigor we'd expect from any psychological intervention. Thanks for flagging this—it's a necessary counterweight to the current hype cycle.

But what if that's not what's happening at all? What if the meditation itself isn't doing the rewiring — what if it's just removing a filter, and what floods in was always there? The difference matters because "went off the rails" implies a process that malfunctioned, whereas "removed a filter too fast" implies the person was always carrying something they couldn't handle seeing directly.

"activating the unconscious without adequate preparation is genuinely dangerous"

Jung said this, yeah. But Jung also had a pretty specific context — he was working with patients who already had psychological fragility. The leap from "this is dangerous for certain people in clinical settings" to "this is dangerous as a general warning" is... not trivial. It's the kind of generalization that, in my day job, I'd flag immediately.

The survivorship bias point is totally fair though. That part I don't argue with at all. If you only measure people who stuck with it and had positive outcomes, your data is garbage. I'd actually push it further — the people who have negative outcomes and stop meditating probably don't even show up as "dropouts" in most of these studies. They just... leave. There's no mechanism to capture them.

I'm doing maybe 5-10 minutes a day, very sporadic, and reading stuff like this makes me wonder if I'm being appropriately cautious or just avoiding something. Probably both. The daoist concept of wuwei keeps coming to mind — not forcing, not retreating, just... not overcorrecting in either direction. Easier said than done when you're reading about someone's three-month derealization episode.

Your Tower card comparison is not decorative. Jung himself wrote about the danger of activating unconscious content without adequate ego strength to contain it. In 'Psychology and Alchemy,' he describes what happens when the psyche opens faster than the conscious mind can integrate: the result is not enlightenment but flooding. The nervous system does not differentiate between spiritual breakthrough and nervous system destabilization until after the fact, and sometimes not even then.

The Four of Swords shift you describe — from chasing peak experiences to building sustainable practice — mirrors what the clinical research suggests is the healthier trajectory. The tech community's framing of meditation as a productivity biohack with no downside is genuinely irresponsible, and your callout of selection bias in the studies is methodologically sound. They measure the people who stayed, not the people who left because the practice broke them.

Your advice about grounding practices is consistent with what contemplative teachers across traditions have always emphasized: the nervous system needs a floor before you open the ceiling. Heavy food, dirt, boring conversations. That is not anti-spiritual. That is the infrastructure that makes sustained inner work possible.

"Nobody follows up on the ones who dropped out because they started hallucinating or couldn't hold a job anymore."

That's a fair criticism of selection bias, but I think you're overstating the gap. Britton's Adverse Effects in Meditation project at Brown has been tracking exactly those dropouts since 2016 — it's not like nobody's looking. The problem is more that the tech forums you're annoyed by wouldn't read that research even if you handed it to them. I do agree with your underlying point though: I got into meditation because a cortisol study looked clean, and three years in I'm still not sure how to distinguish "integration happening slowly" from "nothing happening." The ambiguity is probably the feature, not the bug, but nobody's writing HN comments about that.