I'm astonished you've managed to coax AI into a novel conversation partner. Interacting with AI usually leaves me exactly "super bored and partly drained by the lack of speed and creativity."
All the novelty of AI engagements comes from data it pulls in via tools. It's great to have a conversational interface over the public web, but there's a huge difference between knowledge retrieval and insightful conversation.
Please share how you've cracked this so we <2sd minds can enjoy the crumbs from your lofty table.
This is fascinating. I’m intrigued by the combination of discursive metaphor and weak convention to define actor identities.
The writing style is not natural to me. I was initially put off by phrases like “YAML Jazz” even though I love Tron’s “digital jazz”. Maybe this is because I’ve never experienced psychedelics. I’m at least two generations beyond / behind - removed? - from the movement that defined this way of thinking and communicating.
But sitting with it, it makes sense. I have to stretch my brain. It feels like thinking through a new algorithm… except I’m juggling jello instead of assembling a puzzle.
Ultimately it aligns with my understanding of AI as xkcd’s stirred pile of linear algebra. LLMs don’t understand anything. They project words into probability space, infer momentum vectors based on chunks that bubbling up in the soup of digested everything, and strike out in the prognosticated direction with blind confidence constrained only by temperature.
We constrain these systems with sandboxes and sturdy deterministic tools. Prompts default to the same shape: sharp, precise, hard, directive. But maybe a bag of words is also valid.
If the computer knows your references, each token is much richer than I might assume. Does “minsky” vendor the psychedelic moment? Does “SIMS” scaffold a game loop worth of the emotional energy people have poured into that series? Do these combine to elevate Claude, Qwen, or any other multi-million-dollar number box into an engaging partner?
You got it! The constitutional directive is to lean into the training data, reuse existing terms as much as you can, but when you must mint new terms, they should sound like what they mean.
You can intuitively understand what "yaml jazz" means, and to the LLM, that term is a "k-line" to a skill that spells out the implications: free-form expressive yaml carries structured meaning, and yaml comments add a rich semantic overlay of meaning for both llms and humans, so it follows for both humans and LLMs to write generous narrative comments, and to write deterministic tools (like cli scripts both LLMs and humans can drive) that round trip yaml and preserve comments they can't understand.
The part I'm most delighted with, thanks to its inherent power of simplicity, is the "Selfish" prototype based object system, which supports multiple inheritance from both files and latent space. And it can model other more complex object systems like Smalltalk classes, CLOS MOP, generic dispatch, COM interfaces, OLE dispatch and reflection, LambdaMOO objects, Sims advertisements, etc.
The original Self paper by David Ungar and Randall B. Smith is at least as powerful and mind expanding as great LSD, without any of the downsides! Its ideas underly the fast dynamic JavaScript, Java, C# engines we're all running now, so you can use high doses of it every day for work and for play.
There are some important rules about naming, like "big-endian naming", and "k-lines", a term defined by Minsky, and well represented in the training data because he wrote a lot about them:
>A K-line is a mental structure that, when activated, reactivates an entire constellation of knowledge, skills, and associations. Minsky defined them in a 1979 MIT AI Lab memo, published in Cognitive Science in 1980, then built Society of Mind (1986) on them:
>"When you get an idea and want to remember it, you create a K-line for it. When later activated, the K-line induces a partial mental state resembling the one that created it."
The art is to choose a coherent set of k-lines and terminology, that's well supported and understood in the training data, and fits together well.
If names are pointers into latent space, then naming is API design.
For contrast, Steve Yegge's approach to Gastown is the exact opposite of moollm's approach. This compares Gastown to moollm:
AppleTV does support multiple users, but it’s terrible.
I have an AppleTV on the main television, with Apple accounts set up for each member of my family. It makes almost no difference which user you have selected.
Now, maybe I just haven’t spent enough time trying to set this up, and I’d be delighted to be wrong, but AFAICT:
All users share the same Home Screen layout
All users share the same screen saver
Third party apps have zero integration with the AppleTV accounts. This means I have 4 users on the AppleTV, then 3-5 users on every single app installed on the Apple TV, and any AppleTV user can use any app user.
All together this means the AppleTV users are useless to me. They don’t help me with parental controls. They don’t help me switch between personalized content collections. They don’t even help me log into third party apps!
That is because even though multi-user support exists, the third-party apps with content don't care. They do not map Apple Accounts to their own profiles, or persist profile selection against an active user account.
But many also haven't updated their custom input code to support the current remote design either.
I think some of that is defensive. They don’t have a GPU or other AI Accelerator story right now. They have to remind people that regular compute is still relevant for all the other stuff around the AI workload.
That said AMD’s participation in this round + heavy hints on the podcast and YouTube suggest they will pursue an GPGPU Compute Sled built around an AMD product.
As the person who runs socials, the idea we push AI on socials doesn’t resonate. I get more feedback around “thank goodness you haven’t pivoted to AI-washing your website/marketing”
Well you guys don't talk much on X, which is the center of AI-positive attitudes. I would expect AI-negative response on other social media platforms as they are very prevalent in those spaces, especially on Reddit and Bluesky.
I think I was moreso referring to how I started following some of your employees and then my feed became very ai heavy so I removed them. Like the "nonai users get left behind", "you should be token maxing" type of messaging, etc. Sorry to misconstrue marketing with other forms of PR
totally fair, there's the "Marketing" I'm responsible for and then there's "marketing" which is the broader discussions around something, the algo, etc and I totally respect doing whatever you can to keep your timeline cleaner in a mostly AI-washed hellscape. Know that I am endeavor to do the same with Oxide Marketing.
The related https://fizzbee.ai/ tool is pretty neat. It's similar to /grillme but with additional formalism. Not sure if the resulting specs are definitively better, if only because the FizzBee code is harder for me to decipher.
“On average, each result used three hours of ChatGPT Pro thinking compute with that model. Over the course of the evaluation, the model was posed approximately 4,000 problems.”
That leaves a lot of ambiguity. Is “result” exclusively “positive result shown here” or “all results”?
Since proving infeasibility is a valuable result, the unpublished problems must not have reached a valuable end state. Thus it’s an operator decision of when to turn the machine off and try another problem. I’d thought they had a three hour time box on this, but apparently not.
reply