Have you written about your study technique optimizations anywhere? It would be useful if you could share them.
VlaakithOutrance
Karma: 12
I have found it useful to create a Claude Project called something like “2026 Soul Searching” in which I have added all the context necessary to deal with my personality type and idiosyncrasies. I didn’t want to have all that information in the website-wide “personal preferences” field to avoid Claude always engaging in a personal coach-way. I use it for work a lot and there’s no value there in managing my quirks. The Project serves as the baseline for moments where I know those quirks are a roadblock or have outsized influence on decision making.
One way it has helped engineer change is how strong it is at catching mental patterns before they’re fully on display. I’ll give you an example, replacing the quirk with a placeholder one (not interested in sharing my exact one): Task initiation block. In the Project context, I give full and honest details about how task initiation block has affected my life in recent years, with concrete dated examples, and whether I knew I was doing it at the time.
Because I know I have experienced task initiation block a lot lately, I have developed all sorts of hedging language around decisions that are affected by it. So, if I start a chat in the Project and plainly detail how I’m feeling re: a task/decision ahead, I might feel like I’m nowhere near displaying task initiation block, but Claude is remarkably good at picking up the train of thought that leads me from tnow to tblock. It still catches me off guard.An analogy that comes to mind is when you say “I’m great, how are you?” to a close friend when you’re not feeling great at all, but the words come out with full self-confidence, before the friend snaps you out of it by saying “You don’t sound like you’re great!”.
Intuitively I know that it over-indexes on the context provided in the Project. Sometimes it’ll draw a link between tnow and tblock and I’ll just respond “Nah, I’m good, I’m not displaying task initiation block here”. But when it draws the link correctly, it can snap me out of it, and I’ve had > 10 instances of being helped by the realization that I was slowly moving towards a mental block.
In the conversations I’ve had that fit your definition of Unfolding, there’s been a clear trend where it leads to a realization → realization leads to an exploration of possible solutions → this feels elating and motivating → the next time we connect, the person made no progress and the turmoil unabates.
Making Jake the subject of this trend: his rage comes from a mismatch between his perceived integrity and the lack of agency he’s shown in the face of a tyrant boss → he explores confronting his boss with an ultimatum → this feels motivating because the planning of agentic action feels agentic → the original narrative reasserts itself: “the possibility of speaking it felt out of the question: most likely, his boss wouldn’t get the emotional appeal, he’d be shut down, and things wouldn’t budge.”
I suspect unfolding is augmented by immediately taking action. Conversely, unfolding and subsequently recognizing you haven’t acted on new conclusions can make you worse off: if the picture of your inner battles is sharp enough, repeated inaction makes you feel like a failure. It feels like doing an open book test and still not getting the answer right.I have found that people will be much more likely to act in their best interests (ie acting after thoroughly exploring a problem + its possible solutions) when they take action either during the unfolding conversation or immediately after. It may be because they experience their highest clarity of mind in recent memory; that lightning bolt of lucidity clears up the tangled mess of lies and narratives, and highlights throughlines between A and B. I feel these moments as blank canvases on which I can inscribe the start of a new narrative, which will be strengthened by repeatedly taking action in line with the new narrative. If nothing is done soon after the moment of unfolding, it seems the overarching narrative prevails.
I will give you two examples in recent memory.
Friend A discusses their ongoing mental health battles with me. We cover well-trodden ground, and I commiserate by sharing my own experiences. Like always, this feels good—we feel like we’re getting somewhere as we discuss potential next steps. We reconvene weeks later to find ourselves in the same patterns. I can confidently say we both unfold in the way you describe: our secret justifications peeled back layer by layer until we get to what feels like a bottom. Almost nothing ever comes from these conversations beyond feeling good, which is okay but not always enough. There are only so many ways one can advise a friend to seek professional help.
Last time we chatted, they asked me whether I keep our conversational positivity from slipping away the following morning. I replied I’d only ever achieved progress by acting immediately on the conversational high. Whatever I have put off lately I do immediately after the conversation has ended. It could be messaging someone on the way back home or moving forward with a tough decision in that instant. I take full advantage of the lightning bolt before it fizzles out, in the hope that positive outcomes will carry the clarity further in time. The next morning, Friend A told me they had booked an appointment with a mental health specialist as soon as we parted ways, and that they had gotten done what was previously in the “maybe someday” category for over a year.
Your description of Jake’s situation with his boss is so close to mine that its relatability spurred me to comment. If you remember Jake’s situation, consider mine extremely similar. A few months back, I chose to look elsewhere while still working on the company. Except I haven’t really looked elsewhere in that time, which I justified with a slew of half reasonable arguments. When I “unfold” the whys, I usually feel good in the moment, then do nothing about it.
Cue to me chatting to a family member about this, and they start to unravel my weak arguments. They honed in on personality traits that lead to my avoidant behavior and pointed out opportunities that I wouldn’t touch for unconvincing reasons. They told me I should act immediately. It felt rough in the moment, but also motivating. I also knew I would lose this motivation by next morning, so I immediately reached out to a few people and booked in-person events in the following weeks. I am now thankful I did this, because it would have taken me a long time to do these things separately had I not acted swiftly.
Jake may be at risk of feeling worse if he stops his unfolding short of taking action. Presumably if Jake wants to tell his boss “if you don’t want to do this, I will need to leave” then he is ready to draw a line in the sand. The justification of inaction that follows, whereby “he’d be shut down, and things wouldn’t budge” is a strong sign that he should communicate his feelings the next time he feels the lucidity of unfolding. If the unfolding uncovered integrity and agency as the motivators, their exercise is paramount to Jake closing the loop.
I’d like to publicly take you up on the offer of sharing “any materials you might find helpful for coming to accept/loosen the grip of fear and anxiety”. Do you have recommendations that would likely benefit most people who deal with anxiety?
I would like to know if account-wide custom instructions such as those available on Claude, ChatGPT and Gemini (the only ones I use regularly) override the assumptions made by the model to a prompt, compared with an account that has no custom instructions.
My custom instructions include my job, responsibilities, age and qualifications. But it also includes explicit demands to push back against any claims I make that don’t seem right, an anti-sycophancy barrier. Presumably all this is loaded into every chat’s context so the LLM doesn’t have to bother figuring out my actual expertise level.
The way this shows up in answers is rarely “before I move on to the main paragraph answer, let me push back against your initial claim” or “here’s my main paragraph answer, in which I also push back against your initial claim”. 9 times out of 10, it shows up as a last paragraph called something like “pushback against your initial premise”. But the main body of the answer acts as if I asked the right question with the right terms. Maybe the reason it challenges my assumptions so sheepishly is partly due to the expertise it expects from me due to my custom instructions overriding the assumptions about education.