All of Trevor Hill-Hand's Comments + Replies

"Childish" is the word I also keep coming back to, but hesitate to use, for fear of insulting children by comparing them to people like Sam Altman and Elon Musk.

Reminds me of Solresol, except being a 5-tree makes it so it could fit a pentatonic scale, which is fun.

1Saif Khan
Yesss! Solresol is definitely a spiritual cousin — and you're right, the pentatonic scale connection is super interesting. Kamelo using 5 phonemes intentionally echoes both: * the pentatonic musical scale (so it could be spoken, signed, or played as music), * and tree-based semantic logic, where each level refines the concept more. Solresol mapped syllables to meanings too, but Kamelo’s twist is: * Tree depth encodes specificity. * Context compression is built-in (like pronouns or omitted branches). * Designed from the start to be machine-readable, signable, musical, and tactile. So theoretically: * A blind person could feel a string of raised tactile glyphs. * A deaf person could see it signed. * A device could parse and translate it. * And a musician could sing meaning.

I've actually noticed this in a hobby project, where I have some agents running around a little MOO-like text world and talking to each other. With DeepSeek-R1, just because it's fun to watch them "think" like little characters, I noticed I see this sort of thing a lot (maybe 1-in-5-ish, though there's a LOT of other scaffolding and stuff going on around it which could also be causing weird problems):

<think>
Alright I need to do this very specific thing "A" which I can see in my memories I've been trying to do for a while instead of thing B. I will do thing A, by giving the command "THING A".
</think>

THING B

This is a good point! Typically I start from a clean commit in a fresh chat, to avoid this problem from happening too easily, proceeding through the project in the smallest steps I can get Claude to make. That's what makes the situation feel so strange; it feels just like this problem, but it happens instantly, in Claude's first responses.

I happened to be discussing this in the Discord today. I have a little hobby project that was suddenly making fast progress with 3.7 for the first few days, which was very exciting, but then a few days ago it felt like something changed again and suddenly even the old models are stuck in this weird pattern of like... failing to address the bug, and instead hyper-fixating on adding a bunch of surrounding extra code to handle special cases, or sometimes even simply rewriting the old code and claiming it fixes the bug, and the project is suddenly at a complet... (read more)

5Mis-Understandings
Beware of argument doom spirals. When talking to a person, arguing about the existene of a bug tends not to lead to succesful resolution of the bug. Somebody talked about this on a post a few days ago, about attractor basins, oppositionality, and when AI agents are convinced they are people (rightly or wrongly). You are often better off clearing the context then repeatedly arguing in the same context window. 

I do think it's helpful that managers now have a reliable way to summarize large amounts of comments, instead of making some poor intern with Excel try and figure out "sentiment analysis" to "read" thousands of comments without having to pay for a proper data scientist, and I wonder if that's already had some effects in the world.

Ah, what a fun idea! I wonder if coloring or marking the ropes and/or edges somehow would make it easier to assemble ad hoc- I think Veritaseum's video about non-periodic tilings included some sort of little markers on the edges that helped him orient new tiles, but that was on Penrose tiles and I'm not sure this shape has the same option.

This is absolutely a selfish request, so bear that in mind, but could you include screenshots and/or quotes of all X.com posts, and link to what the post links to when applicable? I have it blocked.

5Shankar Sivarajan
You could consider using xcancel instead.

I thought these were pretty... let's say "exciting"... reads, but I'd be interested to hear more people's opinion of this as a trustworthy source.

I wonder what effect an all-edges pan would have; how did it taste near the edges?

2jefftk
It's not really an edge thing, it's a top vs inside thing. So I wouldn't expect more side surface area to help?

It seems like if there is any non-determinism at all, there's always going to be an unavoidable potential for naughty thoughts, so whatever you call the "AI" must address them as part of its function anyway- either that or there is a deterministic solution?

You can read the fanfiction this is for at: https://www.fanfiction.net/s/14412246/1/Miss-Macross-My-Life-as-The-Star - I'll get around to cross-posting it someday.

All of their work is great, but for my favorite I highly recommend 'Ra', for similar reasons of feeling what it's like to interrogate your own thoughts, senses, and reality itself.

https://qntm.org/ra

Also this fun little story (Valuable Humans in Transit) about an AI: https://qntm.org/transi

I didn't know what to expect, and this was an interesting read. What was the context for when and where it was delivered? EDIT: nm just saw the Fiction tag. Still interested in context though; I do not know who James Windrow is, except for what I can speculate on from this story.

1Alexander Turok
I was imagining a world where creationism was true.

In Anthropic's support page for "I want to opt out of my prompts and results being used for training" they say:

We will not use your Inputs or Outputs to train our models, unless: (1) your conversations are flagged for Trust & Safety review (in which case we may use or analyze them to improve our ability to detect and enforce our Usage Policy, including training models for use by our Trust and Safety team, consistent with Anthropic’s safety mission), or (2) you’ve explicitly reported the materials to us (for example via our feedback mechanisms), or (3

... (read more)

I see people upvoting this, and I think I can see some good insights in this post, but MAN are glowfics obnoxious to read, and this feels really hard to read in a very similar way. I'm sad it is not easier to read.

Something that may help build a better model/intuition is this video from Technology Connections: https://www.youtube.com/watch?v=CGAhWgkKlHI

I mentally visualize the cold air as a liquid when I open the door, or maybe picturing it looking similar to the fog from dry ice.

Since it's cold, it falls downward, "pouring" out onto the floor, and probably does not take more than a few seconds, though I would love to see someone capture it on video with a thermal camera.

After that, I figure it doesn't really matter how long the door is open, until you start talking... (read more)

I don't think it would be TOO long, I happily read through very long posts on here.

However, that said, I was curious enough to read that blog post, and that's about the length and level of detail I expect in a normal short-to-medium size LW post, but it also stopped short of where I wanted it to. I hope that helps calibrate a little? I don't know how "typical" I am as an example LW reader though.

Oh, and because I know it annoys me when people get distracted away from the main question by this sort of stuff, question is "Can you share the experimental resul... (read more)

6bhauth

I've been doing similar things with my day-to-day work like making stuff in CSS/Bootstrap or Excel, and my hobbies like mucking about in Twine or VCV Rack, and have noticed:

  • a similar vibe of there seems to be a "goldilocks prompt narrowness" that gives really good results
  • that goldilocks band is different for different topics
  • plausible-sounding errors sneak in at all levels except the broadest, where it tends more towards very hedged "fluffy" statements like "be careful!"

However, if you treat it almost like a student, and inform it of the errors/conseq... (read more)

I've been doing this for years! When I worked in an office, I had a set of metal chopsticks I was able to leave on my desk — metal was easier to clean.

RE:Footnote #4:

I'll come back to this at some point. Specifically, I'd like clicking that link either to take me to the correct note if it already exists, or CREATE the note if it doesn't exist, while triggering the Templater action that generates all the nice dynamic content on the Daily Note.

I found today, after following this tutorial (which is great, btw, with some tweaks for personal preference this thoroughly fixes everything I felt missing from Obsidian), that putting the template in both the "Daily Notes" template AND as a "Folder Template" mad... (read more)

1Solenoid_Entity
Amazing, thanks!

courage to reject an all powerful authority on moral grounds

This was the most interesting part of the whole story to me, and it's an angle I haven't quite seen in this type of story before. However, I think it was in competition with the personalities of Elohim and Shaitan. They felt too petty and talking-past-each-other to make sense as people from an enlightened race. Maybe if their "conflict" was also a pre-planned part of their strategy, instead of a squabble?

The cultural and literary references didn't bother me, but they did mean that by the end of... (read more)

Oh, and whenever you are able, run things through www.hemingwayapp.com and optimize for shortest length and lowest grade level without losing information.

"I didn't have time to write a short letter, so I wrote a long one instead."

-3Stevarino
Touche.
1Willa
I believe that the Hemingway Editor would be directly useful for my work emails, thank you for sharing :) Sadly, I may not get to use it often for that purpose due to data confidentiality concerns (I work at a hospital).
Answer by Trevor Hill-Hand51

I ended up as part of a team managing the internal communication & knowledge platform for a company that was at the time (early 2020) about ~100,000 employees, now ~146,000. My area of responsibility now includes over 20,000 employees, but I do not directly oversee anyone. I did not have education or much experience particular to this domain, but somehow became a preferred pick for the role, so make of that what you will.

The strategy I've always tried to employ is to treat everyone as intelligent equals, and making as much effort as possible to underst... (read more)

1Willa
I do spend a fair bit of time (a solid business hour a week at least) providing or explaining the correct bureaucratic workflows to customers & even other support staff from different teams. Understanding and communicating those bureaucratic processes is imperative. Knowing when to tear them down / bypass such processes is a good idea, thank you for mentioning it (I have a hard time determining that point). When [where I work] a task or incident reaches such a point, that's usually when escalation to management occurs because I and/or my teammembers have exhausted all our available political options for the small bypasses we're allowed. Management can then run the (what was a task, incident, idea, etc. and turns into the following) problem, change, or project up their chain of command as necessary. Excellent point regarding obtaining everyone's input, and how difficult that can be at times. This is especially true if there's a lack of trust between the individual or group one is interacting with: example, some departments experienced bad support, follow through / communication, and work quality from IT teams in the past, so when I first work with them, there is a necessary stage of trust building that needs to take place before I can meaningfully assist them / perform the work I need to perform. That part (trust building) I've been very good at so far (and received excellent feedback from customers about that). +1 on the insight that most people don't feel silenced (though they may feel ignored in some cases), rather they decide to live with the problem, think it's too small to mention, think it's not important enough, etc. I intend to watch those two YouTube videos this weekend. I'm allowing an educational-exception for those two videos (I'm on a media diet). "Overestimate their intelligence, underestimate their vocabulary." I love that! And would like to add: "..., treat them like a bona fide equal, and listen with deep attentiveness." Note: I ran this commen
1Trevor Hill-Hand
Oh, and whenever you are able, run things through www.hemingwayapp.com and optimize for shortest length and lowest grade level without losing information. "I didn't have time to write a short letter, so I wrote a long one instead."

I think an important aspect to mention explicitly is that it's paired with the phrase "a map that reflects the territory". It's important not because Harold Fiske or the Mississippi River are important to rationality, but because this image exemplifies the idea of that a map is meant to help you understand and reason about something that is not the map.

I agree. Their 'candidate explanations' felt unsatisfying when I got to them, because they spend so much time building up what a good explanation would necessarily feel like. Maybe that was the goal, but if it was, they didn't make it explicit.

Watched this last night. Kurzgesagt is one of the greatest achievements YouTube has enabled, in my opinion.

As a LessWrong reader I had heard a lot of these ideas before, but part that surprised me was Scenario 1: Even if we "only" thrive about as long as other Earth mammals, the 200,000 years modern humans have been around is still only about 1/5th of the way through our story.

I'm doing this as a comment, not an answer, because it's only slightly related to the specific question, but Matt Parker did some videos about similar "impossible" events and/or probability claims, and he includes discussions on why we tend to make errors like that, as humans.

How lucky is TOO lucky? — https://www.youtube.com/watch?v=8Ko3TdPy0TU

How did the 'impossible' Perfect Bridge Deal happen? — https://www.youtube.com/watch?v=s9-b-QJZdVA

I suppose the hope is that then there will be a third tier: "How to move your couch the right way, and why everyone thinks you need avocados for it."

The above comment just helped me realise that the connotation above is why I like the word "credence". Does "credence" have similar problems in other cultures though?

Can you elaborate more on whether there have been noticeable results in either A) taking successful actions based on the most recent predictions or B) improving the forecasting skills of the players? And if so- how were these things measured? How would you prefer to measure them?

3NoSignalNoNoise
We don't have well-defined stats on how well people's prediction skills have improved over time. From my anecdotal observations, pretty much everyone (myself included) starts out vastly overconfident, and then after losing a lot of points in their first few predictions, reaches an appropriate level of confidence. I'm not sure if anyone goes from ok to great though.
3NoSignalNoNoise
There have been few times that we've been able to take actions based on the predictions, because that requires the following combination of factors that tends not to occur together: * The answers are sufficiently clear-cut to make the prediction scorable * There are specific actions that depend on those well-defined answers * Enough people in our local community have enough insight to get some sort of wisdom of crowds. The examples where the predictions led to decisions are: * Scott Alexander visited Boston and we hosted a meetup that he came to, and we wanted to run a survey at the meetup. We took predictions on how many people would attend, along with a conditional prediction market on the survey response rate for paper vs for tablet. Based on this, we went with paper. Attendance was much higher than anticipated, and we ended up running out of forms (but were able to make copies, so it worked out ok). * When our apartment had some maintenance issues and the landlord was giving us mixed signals about whether we'd be able to renew the lease, we took predictions on when/whether the issues would be resolved and whether the landlord would offer a renewal. Based on these, we decided not to look for a new place. The issue was in fact resolved and we were able to renew the lease. * One of our housemates has moved away, with some ambiguity around whether it's temporary permanent, and we have predictions on whether they will return. TBD how this one will go.

Opening in the center, away from you, for sure. Whether symmetric or asymmetric.

In your scenario of carrying in cargo (whether groceries or, say, a sleeping child), having both swing away is a clear advantage.

I think that does make the door significantly less secure against forced entry, but it there's already a secure outer door that's maybe less important?

2jefftk
Yes, I'm not worried about forced entry; we leave this door unlocked and lock the building's exterior doors.

I've been using Roam for about four weeks (found via the Zettelkasten Method article linked in an earlier comment). I wholeheartedly agree with every claim above- Roam lets me freely write down things I want to remember, in a way that I can trust future-me will actually be able to use.

I track commitments using the /TODO feature, and have found that it doesn't even matter when and where I write it (it doesn't even have to be on the page for the relevant project), because all you have to do is browse to the page "TODO" to see all your to-do items.

When you ch

... (read more)

Yes- maybe they're just very uncommon, but I've never been on a bike that I could gear down low enough that it felt easier to pedal than to take a step, on a moderate hill.

8cousin_it
Yeah. Many people say bikes are more efficient at transforming power to movement, so biking should be always easier according to physics, but in reality walking is sometimes easier. I can think of a couple explanations: 1) biking doesn't give the best leverage to your strongest muscles, so you end up tiring out the weaker ones; 2) at slow speed, balancing the bike takes extra effort comparable to walking. I suspect both can be fixed by changing the construction of the bike while still allowing high speed on level roads.

Are there any other Vegas locals that might be interested?

I am actually starting to see this; Droput.tv is one example https://www.thecrimson.com/article/2018/11/19/dropout-tv-review/

But, they still advertise on existing platforms like YouTube, Twitter, etc.

I don't know how that affects this model.

2philh
Another example would be rooster teeth. They have a bunch of stuff on YouTube, but at least some content that's exclusive to their site. (I'm specifically thinking of the latest season of RWBY, I doubt know if there's other examples.)

Wouldn't carrying this analogy to its conclusion mean that you would have to generalize it for a human mind which doesn't already have a bunch of prior understanding of the world? Instead of a universal turing machine, you're using "a universal turing machine that has been taught about witches", which feels like it should be called out as part of the experimental methodology.

So really you would imagine describing the hypothesis and all of the information required to understand the hypotheses to a human that doesn't already kno... (read more)

To be more specific, after rereading the article and thinking for a few minutes, the skill seems to be in correctly deciding whether to accept "everything is a little slow and painful!" as a single big symptom (Mountain), or seeing it as an excuse to not examine and uncover the many small symptoms contributing to that feeling (Cloud). Probably a good place for some heuristics on what bad diagnoses look like.

1Bird Concept
That still seems too vague to be useful. I don't have the slack to do the work of generating good examples myself at the moment.

The difference between Mountains and Clouds seems to be the most critical. They're both described as "problems with many small causes", and now I know they need different strategies, but I don't feel well equipped to notice differences, if any.

2Trevor Hill-Hand
To be more specific, after rereading the article and thinking for a few minutes, the skill seems to be in correctly deciding whether to accept "everything is a little slow and painful!" as a single big symptom (Mountain), or seeing it as an excuse to not examine and uncover the many small symptoms contributing to that feeling (Cloud). Probably a good place for some heuristics on what bad diagnoses look like.

This post helped me notice a difference I've felt between satisfying and unsatisfying explanations; why Feynman explaining something feels different from Wikipedia explaining something. I love it.

There was one famous chicken that was beheaded (during a routine slaughter) just high enough to keep the brain stem intact. A clot coincidentally prevented death from blood loss, and it lived for two years as a touring attraction, before finally dying by choking. It spent most of its time attempting to preen and peck.

http://en.wikipedia.org/wiki/Mike_the_Headless_Chicken