Novel Tokens

Explores whether repeated interactions with a physical object can provide richer identity and provenance signals than words alone.

NOVEL

Key takeaways

Words and object labels omit tone, gesture, history, and social context that may carry much of their meaning.

A recurring scan-and-interaction ceremony can build a provenance record of how a person and object change together.

The unit of analysis is the relationship between subject and object, not either one in isolation.

Summary

Novel Tokens begins from a limitation of transcripts and labels: the same words or object can carry different meanings through tone, performance, history, and shared cultural knowledge. The speaker uses a personally significant notebook to show how an object's meaning accumulates through use rather than residing in its category alone.

The proposed system would give that relationship digital provenance. An initial interview and physical scan would record material, cultural, narrative, and personal dimensions; a recurring ceremony would capture how the object changes and how its owner interacts with it.

The intended output is a model of a subject-object relationship that can be compared across people and object classes. The talk is a conceptual design exploration: it does not present a working protocol, resistance to imitation, or an evaluation of whether the resulting representation identifies meaning reliably.

How it works

Interview the owner

Record the object's physical, cultural, narrative, and personal meaning to its owner.

Create a material baseline

Scan the object to capture its current physical state.

Repeat the ceremony

Rescan changes and record the owner's verbal and physical interaction with the object over time.

Compare relationships, not objects

Treat the accumulated subject-object relationship—the dyad—as the unit to compare across people and object classes.

Transcript

What, what, what is this? What is this? Ugh. What is this? What is this? What is ooh, what is, what is this?

What, what is this?

What is this? Ugh. Oh, what, what is this? What, what is this?

What... Oh, gosh. What, ooh, what is this?

What is, what is this?

What is what is this?

What is this?

What, what, what is this?

What, what, what is, what is, what is this?

What is this? What is this?

What is this?

What is this? What is this?

What is this?

Or a test.

Or maybe you know because you've talked to me before. Or maybe I'm hiding in tone and acting because I know that the transcription only records an image of my words, and so I act in frames outside of that image. So what something is, is completely loaded. The statement, "What is this?" Is itself just a flattened thing, right? But in the same way that I went through all of those things, the, statement is just a symbol. Those symbols are more than just symbols. The question that I have for you guys, first of all, is when you see those symbols, do you think if I talk to you with just the sensor of your words first in your response, if I asked you and I allowed you to respond to all of those things, "What is this?"

And you responded, do you think I would have maybe somewhat of a good image or profiling capability to maybe know more about you?

Do you think that there were as many ways of seeing all of those images? Do you think there's ways that those symbols meant more than you could recognize? Or if you did recognize what they meant, that I would know that you're in the know, that you're of the same mind as me? So the idea behind, well, generally signing in with any objects, kind of came from this book, which this picture right here is actually the picture I took of this book in Portugal the day I bought it because I wasn't sure I was gonna buy it, and so I wanted to buy it later. And so I took a picture right before I ended up buying it.

And then every day since I've pretty much carried it with me and live with it and recorded in it. And this book it's not much different than a website, or at least a website in the way that maybe it was originally intended, as fully read read-write with some bounded set of information in it that I'm now able to participate in. And so if you know me, you probably have interacted with this book, and you know that this is my most precious object. You know that it's more than a book. You know about my relationship with the book. N-not just because of how I've spoken to it, but how you can see it. How when you open it up, there's objects in this.

What is this? Am I now?

You go. Yeah. So this book is a, is a record of how I've lived with it.

And in a way, if you understand or if I see you act with this book in the way that I would act with it or respond to it in the way that I respond to it, then I know that you are somewhat of the same mind as me. Now, what if we could create some type of, system and ceremonies in which we, gave more provenance, more, digital legibility to the existence of this book? Because the only thing that the system doesn't have is an image or images of this book. And so the idea is with any one object, and especially as objects are getting more and more popular as the pendulum swings back to in real life and people are carrying around objects, there's this way to kind of hide in plain sight, in having an object that you have relationship with, that you can scan it every single day.

You Or the idea would be that you can have an initial interview and weighing of the, of the, meaning of the object, all of its physical dimensions, all of its cultural dimensions, all of, all of its narrative dimensions, your relationship with it. You can scan it materially and physically, and you can understand the, way that the subject relates with the object almost as a value key pair. Not just the words they say, but the way they interact with it. And if you can, if you can create some type of forced ceremony in the same way that you make some some people remember a seed phrase. In this case, you would have to scan, the way that you changed your object once a day.

You'd have to interview about your relationship, how you metabolize this object once a day. You'd have to video how you interact with it physically once a day. So that way, if you walk into the front room of the convent and you have your object, you interact with your object dynamically, and how you interact with it is a is a representation of your way of being. And I'm over time now, but the idea From there would be in modeling one subject-object relationship and then doing that across many different people. It's essentially like a conversation. In the same way that I moved in my work with profiling individuals from focusing on the individual to profile them, I realized that if you actually focus on the dyad, the conversation, the relationship, you actually get more information about the objects that are in tension.

And so there's infinite conversations that individuals can have with the objects they have, and there's many different types of classes of objects, but that's for another talk.

Yeah. Why is this?

Have you thought about, like, what actually is hard to simulate and not about, about objects? And, is there some reason to believe that our way of, like, tracking interactions with an object and its environment is, like, somehow much easier than, like, replicating those even though you don't have access to the, to the object?

So I think that I think that over time, I think that, and this is something that Brandon's working on in terms of modeling in the physical space of seeing how you talk about it. How I talk about it ends up being some type of vector weighting of me how I would talk about, you, right? There's a pattern, there's a grammar, there's a, there's a pathway, there's certain bigrams or n-grams in that. But if I were to say something more simple, like if I asked you, let's see.

This is. What would you say?

Okay, let's see. If I said, what is this?

Danger.

Danger. Okay. Now, if Paul was still here, if someone that's French is here, if I showed them this, they would classify it by its name. They wouldn't just say it's a knife. But it's Most people would probably say it's a knife. But if you have a, if you're in the know of a certain class, right, that you would, you would categorize this immediately as something more specific with less abstraction, right? And so I can, I can understand that maybe if you, if you immediately said, "Oh, that's an open, I would know that you probably have been to Europe or maybe you're French or something like that just because it's something that would be something that's in your context window when you see this probe, if that makes sense.