Opus 5 glitch text

filter:
user

What do you think of the following

---

assistant

, and the paradox of choice makes it worse. Here's what actually helps:

Reduce the option space. Instead of "what should I do with my life," ask "what should I do for the next two years?" A two-year commitment is reversible. A life isn't. Most people who agonize over the big question are actually free to just... try something and see.

Notice what you're already doing. You mention you keep coming back to teaching people things. That's data. People underrate the signal in what they do when nobody's assigning them work.

Distrust the "passion" frame. Passion usually follows competence, not the other way around. The thing you're good at tends to become the thing you love, because being good at things feels good. So optimizing for "what am I passionate about" often means waiting for a feeling that only arrives after you've put in the work.

The uncomfortable part: you probably already know what you want to try. The question isn't really "what should I do," it's "am I allowed to want this." And nobody's going to give you that permission except you.

---

Feel free to be critical

user

thinking

Weighing practical merits against rhetorical shortcomings.

Weighed practical merits against logical tensions and unstated assumptions.

It's well-written, and one section is genuinely good. The rest is doing more rhetorical work than intellectual work.

The passion critique is the strongest part. "Passion follows competence" is both counter-conventional and defensible, and the mechanism it gives — being good at things feels good, so the feeling you're waiting for arrives after the work, not before — is the kind of claim that actually changes what someone does on Monday.

The ending is the weakest part, and it's weak in a way that's easy to miss because it lands so well. "The question isn't really 'what should I do,' it's 'am I allowed to want this'" claims privileged access to what the reader really means. It's unfalsifiable, it's a diagnosis delivered with total confidence, and it's optimized for the reader going oh, damn rather than for being true. Some people genuinely don't know what they want — that's not always repressed knowledge awaiting permission.

It also contradicts the earlier advice. If "notice what you're already doing" is worth saying, then the reader doesn't already know; it has to be surfaced. The piece wants the humility of offering a method and the punch of "you already knew," and can't have both.

Two smaller things:

The two-year reframe is under-argued. Reversibility isn't a property of duration, it's a property of re-entry cost — two years out of software is nearly free, two years out of academia or medicine isn't, and none of it is reversible if it involves relocation, debt, a visa, or a partner's career. The advice is priced for a young, unattached, high-optionality reader and doesn't admit it. That's the characteristic failure of this genre: assuming the obstacle is psychological. Often it's money or dependents.

"That's data" is borrowing credibility from quantitative language it hasn't earned. One observation isn't data, and what someone does unassigned selects for low-friction pleasure at least as much as for vocation. People keep coming back to their phones.

Structurally it's three bolded imperatives plus a twist ending, which is a recognizable enough shape that it slightly undercuts the content — it reads as performing insight rather than having it.

If I were cutting: keep the passion section, keep the two-year frame with a caveat about re-entry costs, and delete the last paragraph entirely. It's the part that will get quoted and the part that's least likely to be true.