Where: Kent, near London, UK
When: 19-22 Dec (solstice celebration is Sunday pm)

Come join us for a secular solstice celebration/weekend retreat at Hogwarts School of Alignment and Rationality, Kent - 60-90 min from London by train 

https://luma.com/5n7z4ijw

In the grand tradition of secular solstices (but with less grand ambitions than some), we're hosting a solstice/chill retreat weekend to connect, relax, reflect, process, and of course, celebrate the beauty of rationality and altruism.

Use the Luma page to find out more and sign up! Looking forward to welcoming you at cozy Hogwarts.

ICYMI - Kaj mentioned https://www.lesswrong.com/posts/byewoxJiAfwE6zpep/reality-revealing-and-reality-masking-puzzles in a comment below - I think Anna does a pretty good job of that in that post

I appreciated the work and ideas you shared here. While I understand the kinda natural flow of I designed this cool product > this allowed me to form a clear concept of a potentially useful technique, I think writing 

1. an announcement post for the website and why you think it could be a game changer in collective/democratic decision making and similar fields (with a much more succinct coverage of the object level stuff on SB 1047, given that this is LW; and maybe a bit more yet on the good stuff you anticipate from this kind of product, e.g. in terms of tactical benefits for AI safety); and, separately,
2. a short post going 'here's a neat trick I formulated while working on [link to #1]. I think it could help people get much more out of their discussions'

would have made it easier for your audience to respond better. 

I think you did a pretty good job of conveying why you're enthusiastic both about your project and about your mapping technique, but I suspect
a)  the object-level-induced noisiness of the part where you introduce the website might drown out the core value of it, while
b) the simplicity of the mapping idea may suffer from the curse whereby people think simple ideas were already evident to them just because they make intuitive sense, which might stop them from appreciating the implications/usefulness of the concept. It also seems plausible - this is speculative, obvs - that the LW readership has a distribution that skews towards people who are pretty good at holding a bunch of objects in memory and retrieving them effortlessly, such that maybe they're less excited about this tool than gen pop might lead you to anticipate (I know these last points aren't directly the type of feedback you requested). 

Thanks for this post!

One point I want to add to my previous comment: 

I think it's possible that 'fear of failure + fear of difficulty' doesn't fully explain why people struggle to start/continue an action. 

In fact, I think there's a third reason, one that's fuzzier and therefore perhaps somewhat more delicate to deal with which at least I encounter: I'm often vaguely aware (intuitively/not quite consciously) that a/some second-order consequence(s) will be impacted if I undertake the action, or if I undertake the action before obtaining more information, mustering up sharper focus etc (yes, I realise this is exactly how one might rationalise procrastinating, only about something that isn't the core object). This is esp true if there's a general sense of uncertainty about various ramifications of my executing/prioritising/executing in a particular way the action in question. 

As context, this is something that's become true for me in the course of heavy burnout. Quite simply, I lost confidence in my ability to effectively track ALL the things (hypervigilant? Me? What do you mean, overoptimising isn't optimal?), which caused a bunch of paralysis. I'm working on it; simply becoming aware of these kinds of patterns has helped me tremendously. I think it's worth recognising the impact such mechanisms can have, at least for people who struggle to trust their focus/discernment, either temporarily like me or possibly chronically, if that's a thing for really anxious people. 

I think there's a way this failure mode can be rolled into your 'failure or difficulty' model, by conceptualising this as a subset of the potential failure or difficulty that may arise if you undertake the action.

I appreciate you writing this sequence. Here's why:

- I thought the flow of it was pretty smooth (which is helpful when you struggle with EF, obviously) - I esp enjoyed the fact that you maintained a clear enough map of where you were going at a couple of different levels at basically all times. 
- I got a lot out of your content, mostly in terms of 'cool, I now have an easily accessible and pretty intuitive rough structure in mind of what might be going on when I get stuck and which ways the causal arrows point', which I hadn't taken the time/didn't have the bandwidth to figure out/research myself.
- you made connections with frameworks from adjacent fields and examples that spoke to me effortlessly (relevant level of abstraction, etc) and enriched my toolbox.

Thanks esp for the times you shared your personal experiences, offered a bit of your own speculation (I think?) - and pushed through when you struggled to show up to write.

Quick responses to the questions you asked: 
 

1. What's the % likelihood that you will use AISafety.com within the next 1 year? (Please be brutally honest)
   
   For myself, not super likely as I already have first-hand awareness of a lot of these resources (but I'll likely still be tempted to come to this website to jump straight through a particular URL if I know it's easy to find through AISafety.com vs looking for the exact link in the target website). 
   
   
   1. What list of resources will you use?
      
      Events + training, courses, projects, communities, landscape map (and recommending the reading guide a bunch + likely the donation guide)
       
   2. What could be changed (features, content, design, whatever) to increase that chance?
      
      Really nit-picking here! 
      - I'd find it nicer to use if the menu items were listed at the top of the landing page rather than having to scroll down
      - I'd like the copy font to be more legible (my laptop isn't super bright at the best of time so grey font causes strain)
       
2. What's the % likelihood that you will send AISafety.com to someone within the next 1 year?
   
   I estimate 90%—strong expectation in part due to the wide range of target audiences for whom a referral to the site could be useful + confidence in the diversity of perspectives this hub can send people towards.
   
   1. What could be changed (features, content, design, whatever) to increase that chance?
      
      - This is perhaps going to seem minor but a screenshot of what the Reading Guide landing page view looks like next to the CTA to go read it might give people I recommend the site to a taste of the breezy and interactive feel of the guide. As it is, I can imagine some people might hesitate to click through, maybe expecting a dense 200-page PDF doc that hasn't been updated in a while and then hesitating as to what to do with it. 
      
      - Re intro video: I'd like to see a clickable screenshot of it next to/instead of the CTA. Not the newspaper headline strips, but e.g. the somewhat eerie vibe @ 00'31"
3. Any other general feedback you'd like to share
   
   I have this nagging sense that it might be good to give newcomers a clear, immediate sense of the intentions of the team bringing these resources to people's attention, right at the top of the landing page, basically. 
   As a first-pass suggestion: right under/next to the logo, add 'A volunteer-curated resources hub about the (critical/essential/existential) safety implications of artificial intelligence. About us [<= link]' - or something to that effect :)
    
   Otherwise, I think this is a really valuable initiative and I like the site a lot! I also like that the design conveys a sense of urgency, but one that's grounded in rational concern. Thanks, gang!

Thank you for this post. 

The only thing about which I want to encourage more reflection is Have more "Move fast and break things" attitude [admittedly there is a bit of context I'm not edit-pasting here, but you do seem to favour this approach to a fair extent]. 

My gentle nudge here is based on my sense that 'moving fast and breaking things' can have pretty bad consequences if you're (collaboratively) exploring AI safety research tracks that recklessly put into circulation knowledge that can be used to increase capabilities over/without safety 'points'. 

@author - thank you for this fascinating exploration. I'm appreciating your intuition that there might be some interesting structures there and your ability to point to their shape through this research. It seems helpful to have some sense of the ways LLM may exhibit features/behaviours/geographies that we didn't knowingly 'put there' and that are not easily noticeable. 


@Shayne - Can you expand on 'Literary analysis however probably works dramatically better.  Thus maybe it might be worth looking at the works of literature critics'? 

From what I'm understanding of what your comment is caring for, I don't actually think that the approach you're suggesting here gets you richer/more meaningful findings - structuralism provides its own extremely context-specific (both in time and culture) framing of the symbolic meaning landscape it's concerned with. By comparison, the archetypes you find in the major arcana are pretty steeped in the cultural roots that have underpinned much of the you're proposing to scrutinise instead, and is made up of figures that would have been more foundationally influential in the substrate of today's Western culture (especially with its combination of cosmological influences). Therefore, it seems to me from the way you've phrased your comment that you would propose a shallower tool, essentially, which doesn't strike me as more helpful in making sense of how an LLM might derive a blueprint of human cosmology. 

Also, if the data that shaped the embedding space was all English language text, I think we're lying to ourselves somewhat if we're expecting a diverse, non-Western-centric cosmology to emerge from this kind of analysis. I agree with you that using Lévi-Strauss' frameworks as another track could still be worthwhile, but heeding the previous sentence rather than thinking that the right lens can cure a lack that's actually in the object we're exploring.

Finally, can you clarify what you mean by your very first observation, please?