keltan

• keltanFeb 14, 2025, 11:36 PM

  19 points

  3

  in reply to: Venki’s comment on: Venki’s Shortform

  • Create educational content to sway opinions of large voting demographics. Especially when you can successfully signal that you are a part of that demographic.

  • Do what Zvi is doing but for a lower IQ audience

  • Form organisations in your local area

  • Try your best to avoid making AI a culture war

  • Stay grounded

• keltanFeb 14, 2025, 11:26 PM

  2 points

  0

  on: Bi­modal AI Beliefs

  I think this post points at something quite important. I might suggest adding a TLDR at the top, because the implication section is most valuable, but gets buried.

  Anyhow, strong upvote from me.

• keltanFeb 13, 2025, 9:47 PM

  1 point

  0

  in reply to: Raemon’s comment on: “Think it Faster” worksheet

  Not sure yet. I haven’t had the opportunity to try it out. I expect it’ll be cut down as I use it, for friction avoiding purposes. I’ll update in a week if I’ve found anything interesting.

• keltanFeb 12, 2025, 11:36 PM

  2 points

  0

  in reply to: Towards_Keeperhood’s comment on: “Think it Faster” worksheet

  Haha! I did the exact same thing. I have a Claude project that makes templates for me draft one up. It’s a much longer version of yours. I think I’ll probably steal the one you made instead. But, I’ll post this here in case anyone wants a longer version.

  ---

  type: think-faster created: <% tp.date.now(“YYYY-MM-DD HH:mm”) %> tags: [think-faster, reflection] aliases: [“TF-<% tp.date.now(“YYYYMMDD-HHmm”) %>”] situation: “” duration: 5 # minutes spent on reflection difficulty: 3 # 1-5 scale

  Think it Faster—Quick Reflection

  [!info] Situation What happened? Brief description of the situation or problem that took longer than ideal to solve.

  Part 1: Analysis

  [!note] Actual Steps Taken

  [!tip] Optimal Path What would be the minimum necessary steps to solve this?

  Key Insights

  1. Wasted Motion: What steps could have been skipped?

  2. Missed Opportunities: What clues or approaches did I overlook?

  3. Required Skills: What capabilities would have helped?

  Part 2: Future Application

  [!question] Pattern Recognition Where else might this lesson apply?

  • Past:

  • Future:

  • Next Week:

  Action Plan

  1. Immediate Action: What can I do right now?

  2. Trigger Setup: When should I remember this lesson?

     • When I feel [emotion/​situation]

     • During [activity/​context]

     • Before starting [task type]

  3. Success Prediction: How confident am I this will help? (/10)

     • Confidence:

     • Why:

  Quick Wins

  [!important] Key Takeaway The single most important insight from this reflection:

  ---

  Review Prompts

  • [ ] Did I identify the true bottleneck?

  • [ ] Is my action plan specific enough?

  • [ ] Have I set clear triggers for future application?

  • [ ] Could this insight generalize to other areas?

  [!note] Related Notes Previous similar situations: [[]] Related principles: [[]] Skills to develop: [[]]

• keltanJan 30, 2025, 12:47 AM

  1 point

  0

  on: In­sights from “The Manga Guide to Phys­iol­ogy”

  I just finished up a semester of head and neck anatomy. I went into it for the neuroscience, not expecting much from the other topics. I had a similar experience to you, finding many interesting things that are helpful in my day to day.

  I found the Hyoid bone especially interesting. I remember my first time seeing it on a model skeleton and thinking “hu, someone attached an extra mandible to this guy… and it’s just floating. That’s weird.”

  I then had my mind blown seeing it in a cadaver.

  Side bar: If you ever get the chance to go into an anatomy lab and explore the cadavers I highly recommend. Take what they say about eating and hydrating before going in seriously. ~every 2 weeks someone fainted and banged their head on a metal table.

  The other key takeaway for me was learning about the cranial nerves. Then having the realisation that mine have probably been compromised in some way. (Perhaps COVID?). Which has left me with the humorous and sometimes useful Specific Anosmia of not being able to smell farts.

  —

  Reading the Manga Guide to statistics has been great for Anki cards. Easy image occlusion cards that make reviewing more fun.

  —

  I wish I liked cells at work more. Still, I was able to explain where blood comes from to a kid yesterday because of things I learned in that show.

  —

  We seem to have overlapping interests. So I’d like to recommend “Curious” by Lily Serna.

  Perhaps it isn’t applicable to you. But it took me a day to read, and I added 100s of Anki cards derived from it.

  Lily is famous for doing mental math super quickly, and the final section of the book is just a bunch of the tricks she uses clearly written out. With a bit of thought, these tricks combine into more powerful mental movements that have sped up my mental calculations quite a bit.

  The rest of the book is cool math trivia and life hacks. Very fun, a bit simplistic.

  —

  I’m enjoying these media reviews, excited for the next one!

• keltanJan 15, 2025, 10:22 PM

  2 points

  0

  in reply to: sapphire’s comment on: We prob­a­bly won’t just play sta­tus games with each other af­ter AGI

  Sex is fun and awesome. Though it doesn’t feel fun and awesome to have sex all day everyday. You could probably do transhuman meth and make sex fun all the time. But a Pleasure Cube/​Super Happy scenario makes me sad.

  I’m also wondering who you’re talking about when you say “most people” here? I have the opposite model of most people.

• keltanJan 15, 2025, 2:37 AM

  1 point

  −1

  on: How do you deal w/​ Su­per Stim­uli?

  //​This Comment contains references to self harm.

  This idea is a little crazy, but… You could use a TAP that punishes you every time you open the app.

  Trigger: I just clicked the Twitter Icon

  Action: Bite my arm until it hurts

  Biting is an example, But you could also:

  • Have your phone auto-play fingernails on a chalkboard sound

  • Slap yourself

  • Punch your knee

  • Just have the action be closing the app immediately

  • Flick yourself in the nose

  I’d recommend reading the Hammer Time Sequence post on TAPS. Specifically, the part about setting a Yoda Timer and practicing it for 5 minutes.

[Question] Where should one post to get into the train­ing data?

• keltanJan 14, 2025, 9:31 PM

  4 points

  0

  on: How do you deal w/​ Su­per Stim­uli?

  For those who live alone, one option for the phone password is to make it an Antimeme.

  1. Randomly generate a long random string

  2. Mix in some novel Unicode characters that you’ll have to remember the names of so that you can google

  Write it down somewhere inconvenient

• keltanJan 8, 2025, 2:03 AM

  2 points

  0

  in reply to: CstineSublime’s comment on: quet­zal_rain­bow’s Shortform

  I’ve been thinking about this a lot lately. It seems to link to many things. And might be a bit too much for just a comment. But here are some key concepts from mostly psych that I think link to why sleeping on a problem makes it easier.

  • Hebb’s Law

  • Learning is assumed to take place over a 24hr span

  • Chunking

  • The Multi-component Model of Working Memory

  • Mice developing ‘Maze Neurons’ when learning a maze

  • People who are woken mid-sleep and self report dreaming about a problem they’ve tried to solve, do better the next day than people who are woken and don’t report dreaming about the problem

  If I boil it down, I have two hypotheses that could both be true.

  1. When you dream about a problem you’re brain is formulating ideas that can help you solve it. All you have to do the next day is try again and those ideas will become available to you as if you had just ‘had an idea’

  2. Sleeping on a problem breaks it up into more manageable chunks that you can better manipulate in working memory the next time you try to solve it.

  There are other things that happen during sleep that will just make every problem easier to solve the next day. For example:

  • Cleaning up chemical ‘garbage’ that collects in your brain during the day.

  • Forgetting things that the brain doesn’t think you have a use for

  • Resetting/​reducing your emotions. (If you’re stressed about a new problem, you’ll find it easier to solve it when you’re less stressed.)

• keltanJan 5, 2025, 4:36 AM

  3 points

  0

  in reply to: quetzal_rainbow’s comment on: quet­zal_rain­bow’s Shortform

  Hard agree. I think sleeping on a problem is underrated. But even though I think that, I still fall into the failure of “I don’t get it. I must be dumb or something”.

• keltanJan 4, 2025, 10:39 PM

  3 points

  2

  on: [Fic­tion] A Dis­ney­land Without Children

  While many of the review requirements aren’t applicable to this writing. It doesn’t lessen the impact it has.

  This is a horror I would like to avoid. I think Sci-fi of this sort helps to prevent that future. This is something my non-technical Mother could understand. Something I could show people to explain the worst.

  I will think of this post as the future goes on. I am desperately trying to make this story one that we look back on and laugh at. “What silly worries” we’ll say. “How naive.”

• keltanJan 3, 2025, 5:51 AM

  5 points

  0

  on: keltan’s Shortform

  A LW LLM I would like is a “What post was the one where Eliezer talks about logic or maybe algebra or something? I think it might have been in Plane Crash but idk?”

  Sometime I hit a roadblock in my thinking because I’m trying to remember a specific LW post that talks about a skill I’m currently trying to practice. It’s quite flow breaking to try and search the normal way. Current LLMs aren’t up to the task and Hallucinate LW posts every time I do it.

• keltanJan 2, 2025, 10:36 PM

  1 point

  0

  in reply to: CstineSublime’s comment on: keltan’s Shortform

  Closer to the first one. I find when writing to think my mind has two modes. Very system 1 and system 2. If I’ve been going for a while on a side branch system 1 takes over. The writing becomes less about thinking and more about the act of writing. This leads to me making a hypothesis and saying “idk why that is”. That triggers the alias, which points out to me that I’m not really ‘thinking’. I then switch to “How could I test if this is true?”

  I appreciate the 25% reframing. That’s something I wish I’d thought faster.

  Trigger: I see a %# Action: Switch it in my head

• keltanJan 2, 2025, 3:25 AM

  2 points

  0

  in reply to: CstineSublime’s comment on: keltan’s Shortform

  This isn’t an extremely useful technique. What it really does for me is break me out of undirected thinking with my writing and get me to actively start thinking things like “ok, but why would this be happening?”

  I think 75% of the time it’s not helpful. Sometimes unhelpful when breaking a flow.

  I’m working on thinking things faster. Though, it’s not a skill I’d say I have yet.

  However, it’s pretty low cost for any payoff at all.

• keltanJan 2, 2025, 12:33 AM

  2 points

  0

  on: keltan’s Shortform

  In my Obsidian.md vault I use the note aliases to point out to myself when I’ve said something that indicates that I Am Confused. I’ve progressively added and pruned this over a year. Here are the “Confusion Phrases” that I currently use.

  ---

  aliases: I am confused, I don’t understand, This is confusing, This is very confusing, I am quite confused, Wait what, I don’t know where I fucked up, I Feel Confused, I notice I am Confused, I’m confused, I don’t get it, I just don’t get it, How am I meant to, How do I understand this, That’s surprising, I’m not sure I understand, I’m completely lost, I’m lost, you lost me, I have no idea what’s going on, what are they even trying to say, what are they trying to say, what is she trying to say, what is he trying to say, I’m lost for words, I can’t wrap my head around this, I can’t get my head around, I’m struggling to make sense of, I guess I just forgot about, IDK, I don’t get,

  ---

  Kinda related: I was inspired by the Fooming Shoggoths to make a song based on the original Your Strength as a Rationalist post a while ago. Here’s that song.

• keltanDec 29, 2024, 10:56 PM

  1 point

  0

  in reply to: abstractapplic’s comment on: Does Claude Pri­ori­tize Some Prompt In­put Chan­nels Over Others?

  Much appreciated!

  I’ll:

  • Correct that typo

  • Add a section back in to hopefully make it less confusing

Does Claude Pri­ori­tize Some Prompt In­put Chan­nels Over Others?

• keltanDec 29, 2024, 12:24 AM

  3 points

  2

  on: AI Psychology

  I decided to create this tag for two reasons:

  1. The concept of LLM Psychology is interesting and exciting to me

  2. I have progressively seen more people referring to a type of research as LLM Psychology. I think having a place specifically for it on LW is useful.

  If you reply to this comment with posts you think fit under this tag, I’ll read them and decide if they seem like they should be here. I’m currently quite fuzzy on what really belongs in this tag. Clarification on what you think LLM Psych is would be much appreciated.

• keltanDec 22, 2024, 1:34 AM

  4 points

  0

  on: keltan’s Shortform

  My timelines have now updated to something closer to fast takeoff. In a world like this, how valuable is educating the general public? Claude claims science started worrying about the climate in the 50s/​60s. It wasn’t until 2010s that we saw meaningful action beginning to take place. Do we have the time to educate?

  To be clear, this is more of a question than an opinion that I hold. I am working to form an opinion.