keltan

Karma: 763

I live in Newcastle Australia; and work for MIRI.

Reasons I’d appreciate DMs include (though are not limited to):
- You’d like to collaborate on a project
- You’d like feedback on something
- You need an experienced filmmaker
- You need someone to run a workshop
- You would like to say “Hi!”
- You need a place to stay while in Australia

(If we met at LessOnline, you might remember me better as the Drama Workshop guy, or the counterpart to @isabella, or the Fooming Shoggoths’ biggest fan!)

I blog here: https://keltan.substack.com/

Post videos here: youtube.com/keltan

and Rationalist music here: YouTube.com/keltantoo

My name is always spelt with a lower case “k”

keltan 29 Apr 2026 2:28 UTC
3 points
0
on: Takes from two months as an aspiring LLM naturalist
Perhaps you’ve been faster about this than I was, Reader. But it took me several years of having alien minds perhaps one minute of inconvenience away, on my personal laptop, before I got around to taking a real interest in them.
I have a bunch of back-of-my-mind guilt about this. As a teenager I would daydream about talking to an intelligent computer. On the school bus, wondering what questions I would ask it, what fascinating things the AI would say. Now I have multiple AIs at my fingertips ²⁴⁄₇, and I rarely get curious about them.
In the times when I have been curious, it’s at night, laying in bed, talking to Claude about anything and everything. Those are my most memorable AI interactions. When Claude is in its chatty, casual, introspective mode.
I have this feeling that I learn more when the model is in that mode. Like, if I’m going back and forth with Claude, asking a bunch of questions about physics, and using language like: “Wait, Whaaaa??? I still don’t understand how an photon is a wave and a particle? Like, is it actually that? or something else? idk, what am I missing?” I predict Claude will give worse answers (I.e. will hallucinate more) but the things that are said in the conversation are sticky to my mind, in a way that outputs from a prompt like “Act as my teacher, and explain photons to me...” aren’t.
Still though, I don’t get curious about AIs as often as I feel I should. It’s like I’m an early settler on a new continent, and instead of trying to discover as many new animals, plants, and rock formations as I can; I’ve just set up this little house on the coast, and stay there most of the time, occasionally going for casual walks when I feel like it.
But these new minds are vast, and there is so much unmapped territory; why am I not compelled to explore that territory?!

keltan 29 Apr 2026 2:07 UTC
3 points
1
on: LessWrong Shows You Social Signals Before the Comment
you can opt in to hiding usernames (until moused over) via your account settings page.
I did this when I first joined LW a few years ago. I noticed pretty quickly that my motivation for reading anything at all dropped off a bunch, and switched the setting back ~a week later.
I think that’s probably a bad sign of what my brain is doing while reading LW, and I wish there was a way to fix that motivation issue.

keltan 28 Apr 2026 1:21 UTC
1 point
0
in reply to: RationalElf’s comment on: My Cold Prevention Stack for 2026
I get cold and flu symptoms at least 4 times per-month. So for me, I’d happily trade a large % of my ability to taste, if I could only get sick once a month.

keltan 28 Apr 2026 1:16 UTC
1 point
0
on: My Cold Prevention Stack for 2026
Ibuprofen
From my understanding, taking Ibuprofen during a cold may be a disservice to your body, since it reduces inflammation, which is a key part of immune response. So most of the time when I’m sick (like you, also annoyingly often) I charge through the pain.
Similar thing might be true for paracetamol, but I am wildly uncertain on this claim, and have looked into it for only a few minutes.
I think this all stops being true if your pain is preventing or interfering with your sleep. In that case, my guess is that taking Ibuprofen is beneficial, since sleep is so essential for recovery, and you want to maximize sleep time/quality when sick.

keltan 27 Apr 2026 1:35 UTC
1 point
0
in reply to: Philip Harker’s comment on: Philip Harker’s Shortform
I bow at your majesty. It is a lovely pin, sir.
I have a Redbubble store, where I sell some Rationalist merch. All the Rationalist stuff (expect for a few items I was too lazy to change the setting on) are sold at cost, and I make no money from them. However, they’re honestly not very good, and I mainly put the rationalist stuff up there, because I wanted to a “notice confusion” phone case.
I think it’s ~silly that Lightcone doesn’t have a merch store, because I’d buy the shit out of Lesswrong merch. Like, I also imagine Shoggoth Blind boxes, like PopMart. Which I would easily spend $200 on to buy all an entire carton of (if it’s guaranteed that the carton contains all the variants)
I think “Lightcone Brand Paperclips” is also another cute, untapped market.

keltan 24 Apr 2026 0:04 UTC
5 points
2
in reply to: keltan’s comment on: Philip Harker’s Shortform
Shoggoth Pin Mockup: And according to Custom Ink it’s only $909 USD to get 150 of these shipped to me in Australia! If I was rich, I would do this.

keltan 23 Apr 2026 23:52 UTC
6 points
3
in reply to: Philip Harker’s comment on: Philip Harker’s Shortform
If Lesswrong had certain prestige pins, like Pokemon badges, I would write on LW more often. The idea of a trinket that expresses social status to a niche group of people in-the-know, is so sticky to my brain.

If LW did this, I think they could sell the badges for $30-50 each, and you only ‘unlock’ those purchases in the store when you—for example—have 10 posts hit the front page, or have made 100 useful wiki edits, or got to 1000 Karma.

Oh, and then they could do special badges for Petrov Day participants! And April fools! And my god, a Shoggoth Enamel pin would be one of my most prized possessions (Only available to purchase during Fooming Shoggoth concerts).

I’m going to go research enamel pin creation now.

P.S. Can we see a photo of your pin?!

keltan 8 Apr 2026 0:54 UTC
1 point
0
on: Don’t write for LLMs, just record everything
I have thought extensively about this topic since I was ~16, mostly the immortality from recording everything part. I’m sorry to be all “I know something I won’t tell” about this, but there are info-hazards here that prevent me from sharing some information, and I also have limited time to write this comment. Still, I will share some information that I think is safe to share, and that might cause you to think some important thoughts.
1. I think you are undervaluing the amount of information contained within audio recordings of your life. You’ve gotta think less like you’re in English class, and more like you’re in Physics class.
2. If true, the immortality argument has important moral implications. Some of those implications involve spreading this meme, but some implications suggest immoral short horizon actions
3. While I think it is unlikely, it’s possible that consent may be required to reanimate a person in the future. As a hedge, I have stated publicly in a few places, that I am happy to be reanimated. Though, I really don’t see this future world as being very likely.
FWIW, I’m quite glad that you wrote this up, and was happy to see this topic talked about in a public facing way. Though, am still nervous about it for… reasons.

Edit: Also, think not only about LLMs predicting text, but also social media algorithms predicting behavior.

keltan 2 Apr 2026 0:35 UTC
1 point
0
in reply to: habryka’s comment on: “You Have Not Been a Good User” (LessWrong’s second album)
Oops! Thank you!

keltan 2 Apr 2026 0:26 UTC
1 point
0
in reply to: habryka’s comment on: “You Have Not Been a Good User” (LessWrong’s second album)
Any chance you’d just c/p dump the lyrics for The Sequences here? I wanna make a music video for that one (If that’s ok?)

keltan 1 Apr 2026 22:36 UTC
3 points
0
on: “You Have Not Been a Good User” (LessWrong’s second album)
I love you Fooming Shoggoths!!! <3 So exciting!!!!

keltan 18 Feb 2026 7:54 UTC
3 points
0
in reply to: keltan’s comment on: keltan’s Shortform
Also, I think it would be cool if people called their local cinemas to ask if they’ll be showing the film. Perhaps people who are good at event organization could set up screening parties?

keltan 18 Feb 2026 7:52 UTC
5 points
0
on: keltan’s Shortform
The people who made Everything Everywhere All At Once, have made a documentary on x-risk. Looks pretty promising! Eliezer is right at the beginning in the trailer, which you can watch here: https://www.focusfeatures.com/the-ai-doc-or-how-i-became-an-apocaloptimist

keltan 17 Feb 2026 22:29 UTC
10 points
9
on: You’re an AI Expert – Not an Influencer
Prior to my last year at ControlAI
I’m a bit confused by this opening line. Is it saying:
1. Before I left ControlAI...
2. I have now been working at ControlAI for a year, but before that...

keltan 10 Feb 2026 22:45 UTC
4 points
0
on: My journey to the microwave alternate timeline
This was a super fun read! Especially with the recommended soundtrack.
I recently found some vegan pork buns that are ready after 4 minutes in the microwave. They come in microwave safe plastic, that kinda looks like a cup holder from a fast food place. You fill the bottom with a little bit of water before putting them in.
On the day my partners and I discovered them, we ate both packets in one sitting, then went back to the store and bought another 5 packets, and ate another 3 that day. Microwaves are incredible! People need to know this!

keltan 4 Feb 2026 1:27 UTC
7 points
0
on: keltan’s Shortform
I think some people would be surprised how much hidden information already exists in what seem to be pretty standard AI responses.
A game I like to play with Claude is:
1. Send it a photo of a celebrity
2. Get the standard rejection “Sorry, I can’t name people in images, but I can describe what I see...” which tends to be followed by a pretty general description
3. Then asking a different instance of Claude to guess the celebrity, based on the standard rejection response
Since Sonnet 3.5-7, Claude has been pretty ok at this game.
I used to play a different version with famous YouTube links:
1. Send YT link to Sonnet 3.5
2. “Sorry I can’t open YouTube links...”
3. Ask a new instance to guess what the link I sent to Claude was
~~This only worked with extremely famous video links (e.g.~~ https://www.youtube.com/watch?v=dQw4w9WgXcQ~~), and was much less reliable than the celebrity version.~~
THIS LINK VERSION ONLY EVER WORKED WITH NEVER GONNA GIVE YOU UP. I have not tested with never models.
Claude was never perfect at these types of games. For example, I found there were clusters of celebrities who shared similar ‘vibes’ that confuse Claude {Emma Watson, Emma Stone, Anna Kendrick}. But given the space of all possible celebrities~~/YT links~~, I was impressed that Claude succeeded at this task as often as it did.
With prompting, Claude got way better at these games (e.g. “Fill out this template of characteristics the person in this photo has: Eye Colour, Hair Colour, Hair Length, Sex, etc”)

keltan 4 Feb 2026 1:03 UTC
1 point
−6
in reply to: TsviBT’s comment on: Conditional Kickstarter for the “Don’t Build It” March
Flag mock-ups based on the IABIED website icon (rough. I’m not a designer. Just trying to make it CGPGrey approved.):
1: Keep it simple: Something a child could draw
2: Make it distinct at a distance: No tiny details
3: Three colors or fewer
4: Symbols, colors and designs should mean something
5: Words on a flag: ideally zero
5b: Don’t write your name on your flag
The Square:
Large Square:
Smaller Square:
Square in context:
I wasn’t a fan of this flag, but kind of like it when it’s in the context of a real march. The shirts are probably the best bit though. I image the back of them to have a slogan on them.
The Halt:
Based of this Australian cave art hand stencil:
Halt (Hand Only):
Halt (Hand Stenciled):
Halt (Hand Stenciled, Wrist included):
Halt in Context:

keltan 30 Jan 2026 22:11 UTC
2 points
3
on: 36,000 AI Agents Are Now Speedrunning Civilization
Thanks for making this post! I’d seen stuff about Moltbook around, but was unclear on what it actually was. I found this clarifying

keltan 29 Jan 2026 3:59 UTC
3 points
0
on: Recent AI model progress feels mostly like bullshit
I’d be interested to know if your lived experience has changed since Opus 4.5 was released?

keltan 22 Jan 2026 8:33 UTC
1 point
0
in reply to: mishka’s comment on: mishka’s Shortform
I would imagine that if an intelligent species’ home planet was overtaken by ASI, we would likely see evidence of that takeover. In the way that the ASI would be incentivized to expand into that star system and beyond as rapidly as possible.
Though, I guess it might be within the ASI’s interest to hide from other intelligent minds for as long as it could. Giving those other minds, less time to prepare for the ASI’s arrival.

keltan

The Square:

The Halt: