Category: Other Page 20 of 177

Multi-Concept Addressing for Latent Space Navigation

On 18 September 2024

Preface: How I got here

I’ve been hanging out at the library lately, and realized that my Dewey Decimal System (DDS) knowledge has gotten pretty rusty. I found a “concise” summary of it here, and printed off the First, Second, and Third Summaries, which cover The Ten Main Classes, The Hundred Divisions, and the Thousand Sections, respectively.

I knew there was some controversy about the DDS, but I hadn’t checked in on what it was these past couple decades, I guess. But in skimming through the classes, divisions, and sections, it became apparent how lop-sided its distribution of identifying numbers is towards all things European. We see it again and again that it literally marginalizes entire cultures and their achievements by sticking them into “grab bag” left-overs like:

290 Other religions (where 220-280 are all overtly dedicated to Christianity, and 200-219 are no doubt heavily influenced by that tradition)
490 Other languages (420-4880 are all European languages)
890 Other literatures (810-880 are all Euro or American lit)

I won’t bore the non-taxonomically inclined among you by going line by line through The Thousand Sections (though I am strongly tempted to, but that would prolong this preface unnecessarily moreso than already). But some curiosities jump out in that Islam, 297 doesn’t have its own number to itself, but also includes Babaism & Bahai Faith. Likewise, Buddhism is not given its own name let alone its own number. It is a subdivision of 294, Religions of Indic Origin. I don’t know the exact numbers, but based on some initial skimming on Perplexity, it looks like Buddhism + Islam combined in terms of number of followers on Earth is roughly equivalent to that of Christianity globally. But the Dewey Decimal System doesn’t represent these other dimensions of social reality.

Anyway, all that is to say, with an eye to not duplicating the failings of the DDS as a metaphorical jumping off point, couldn’t it be an interesting exercise to come up with some kind of flexible, less judgemental addressing system for navigating high-dimensional latent spaces, such as those you encounter as a user of generative AI models and systems? I’ve already experimented in this direction visually in the past, thinking about how gen AI image creation systems like Midjourney could benefit from some kind of hand-held controller, which would let you rapidly assign and modify dimensional values on the fly, in order to traverse neighborhoods and relations in more or less real time. Latent space as navigable VR, if you will.

I took this problem to three different AI genies, of, paraphrased, give me a dewey decimal system for latent space. The first genie, whose angelic name is ChatGPT-4o, gave me answers that were mildly insightful, but not adequately interesting to pursue in depth. The second genie, whose moniker is Claude, gave me results which were promising, and a UX response that was riddled with errors and hobbled by rate limits. The third genie, named for the Winds brought with it clear thinking, and an in-depth ability to solve the problem through interrogation. That genie’s failing is its refusal to follow custom instructions or the equivalent at a prompt level (“code only, no explanation”), and it’s slow speed. But what it lacked in those areas, it made up for in its ability to guide me to towards a tentatitvely adequate V1/MVP, which is presented here without further ado after this absurdly long preface. My sincerest sorry/not sorry.

Introduction

Disclaimer:

The Multi-Concept Addressing system (MCA) is an attempt by a non-technical author to develop a preliminary schema for one way of potentially addressing locations within latent spaces. It may not prove to be the “best way,” but seemed good enough to at least put out to get the conversation started.

Much of the rest of this text that follows comes directly from Mistral, with light edits from myself.

MCA: The “Dewey Decimal System” for Latent Space in Generative AI

Multi-Concept Addressing, or MCA, is a proposed addressing system designed to navigate and interpret high-dimensional latent spaces in generative AI models. It provides a structured and interpretable way to represent complex scenes and images, much like the Dewey Decimal System organizes information in libraries. (*See: Preface)

Key Components:

Base Concepts: High-level concepts that define the broad categories of elements in a scene.
Sub-Concepts: Detailed information about specific elements within the base concepts.
Relations: Relationships between different concepts, capturing how they interact.
Context: Additional contextual information that provides nuance and depth to the scene.

Operational Principles: MCA operates on membership degrees or intensities (that is, whether an image, for example, contains members of a particular concept, and how much), allowing for precise control over the presence and importance of various concepts and relationships within a scene.

Problem Solved: MCA addresses the challenge of navigating and understanding high-dimensional latent spaces in generative AI models in something that approximates a human-readable format. It provides (hopefully) a holistic and flexible solution that can potentially be adapted to various contexts, including image generation, semantic analysis, and data retrieval.

High-Level Example

Consider the following natural language prompt:

a cat riding a bicycle wearing a football helmet playing a banjo in outer space

This prompt contains multiple concepts and relationships that need to be represented in a structured and interpretable way. Let’s see how MCA can achieve this. [Back to my text with Mistral excerpts included below.]

The first part of an MCA address consists of a string like this representing base concepts and weights:

Ani90Obj80Env90Act80Acc70

Where the name values for high-level (or “base”) concepts represented in this query are:

Ani: Animals
Obj: Objects
Env: Environments
Act: Activities
Acc: Accessories

Using only this for addressing just leaves us in a very fuzzy general vicinity… maybe something like a room in a given library, or a big shelving unit. We might be able to find what we need, but we’re most likely to stumble around looking for it without more specific information.

Base Concepts Sample List (Provisional)

As an aside, I had Mistral work up a set of what might be the top base concept names and abbreviations. I kept saying, do you have any more edge cases, and it kept giving more and more. Eventually I gave up, as this seems like an adequatedly representative step for a v1 of this concept. Here is that list, for completeness (though it also made a list that was much much much longer, and I had to eventually push the stop button. I’ll at least spare you that one). I think there could be better three-letter codes representing each concept, but I left them as the first three letters to make it simple. Here it is:

Animals (Ani)
Objects (Obj)
Environments (Env)
Activities (Act)
Accessories (Acc)
People (Peo)
Plants (Pla)
Structures (Str)
Weather (Wea)
Time (Tim)
Emotions (Emo)
Events (Eve)
Sounds (Sou)
Text (Tex)
Abstract Concepts (Abs)
Technology (Tec)
Food and Drink (Foo)
Transportation (Tra)
Art and Culture (Art)
Natural Phenomena (Nat)
Science and Mathematics (Sci)
Health and Medicine (Hea)
Education (Edu)
Sports (Spo)
Mythology and Folklore (Myth)
Fantasy and Science Fiction (Fan)
Geography (Geo)
History (His)
Lighting (Lig)
Colors (Col)
Textures (Tex)
Movement (Mov)
Interactions (Int)
Symbols (Sym)
Virtual and Digital (Vir)
Celestial Bodies (Cel)
Microorganisms (Mic)
Chemicals (Che)
…and on and on

Navigating Sub-Concepts

Getting back to the addressing for that specific reference prompt results – if we’re looking at all of this in JSON, then the next part of the address will derive from sub-concepts within those broader base concepts which could be represented like this:

{
  "SubConcepts": {
    "Ani": "cat90",
    "Obj": "bike80banjo70",
    "Env": "space90",
    "Act": "ride80play70wear60",
    "Acc": "helm70"
  }
}

So our MCA partial address now has the base concepts, and sub-concepts, which would look something like this, give or take:

Ani90Obj80Env90Act80Acc70::cat90bike80banjo70space90ride80play70wear60helm70

It looks inscrutable-ish, but it’s not really. It’s just a way of compressing the JSON schema into a single line.

Relations & Context

But still in its current form, we don’t necessarily know enough about the different entities and actions to find exactly what we’re looking for. We might get close, but still have major errors. We need to know something more about the relationships between all these entities or values that are named, as well as any larger context not otherwise captured in the address so far. Otherwise our accuracy is going to be pretty low for navigation.

In JSON, these might look like:

  "Relations": [
    "Cat-RidingOn-Bike",
    "Cat-Wearing-Helm",
    "Cat-Playing-Banjo"
  ],
  "Context": [
    "Surreal",
    "Humorous",
    "BrightColors"
  ]

And if we squish that back down into the full MCA multi-concept address, with weights, it might look something like this:

Ani90Obj80Env90Act80Acc70::cat90bike80banjo70space90ride80play70wear60helm70::Cat-RidingOn-Bike_Cat-Wearing-Helm_Cat-Playing-Banjo::Surreal_Humorous_BrightColors

Granted, we could probably also assign weights to the relations and context elements, but I didn’t want to complicate it any more than it already is.

Putting It Together: Full MCA Example Schema

Putting the above extended MCA address all together again as JSON to elaborate its components:

{
  "MCA": "Ani90Obj80Env90Act80Acc70::cat90bike80banjo70space90ride80play70wear60helm70::Cat-RidingOn-Bike_Cat-Wearing-Helm_Cat-Playing-Banjo::Surreal_Humorous_BrightColors",
  "BaseConcepts": "Ani90Obj80Env90Act80Acc70",
  "SubConcepts": {
    "Ani": "cat90",
    "Obj": "bike80banjo70",
    "Env": "space90",
    "Act": "ride80play70wear60",
    "Acc": "helm70"
  },
  "Relations": [
    "Cat-RidingOn-Bike",
    "Cat-Wearing-Helm",
    "Cat-Playing-Banjo"
  ],
  "Context": [
    "Surreal",
    "Humorous",
    "BrightColors"
  ]
}

Is this actually simple and flexible like I had hoped setting out on Today’s AI Side-Quest™? Hard for me to be the judge, but so far it is the only thing of its kind that I have found out there (though I did find some adjacent concepts I’m not yet well-versed enough in these areas to explore in depth), and it does to me at least address a real and specific need, whether or not it, erm, completely accurately gives a reproducible address every time in all situations. It is still maybe a bit vague, but at least hopefully narrows down the task of navigation into a more restrained dimensional space, with keys as to values that could be changed in searching for the specific “shelf” that contains what you are after.

Really, what I imagine in all of this is like a bunch of conceptual characteristics mapped to sliders in a UI, where fully on means that a given characteristic/dimension/concept/tag is applied to the max within the desired outputs. And fully off means that attribute is excluded, plus all the values between. Then using a machine mapped to this would be about playing around with different conceptual sliders to emphasize or de-emphasize members of a given group or groups of high-dimensional characteristics in the latent space.

Phew, lot of words to get out here, but I think that brings us to the end for now, if not the “conclusion.”

Post Script

I am trying to find a natural language UI design tool that can output a version of the above as a simple web app, something to the effect of what I described in this prompt:

app for navigating addresses in latent space based on given values (concepts, subconcepts, relations, context) and their weights. the app consists of sliders paired to specific example attributes or concepts which can be adjusted to yield different results in a viewer window that shows that location

This is somewhat janky, but a quick version of that made using UIzard.io just to leave you with something more concrete to consider:

Adapting Open World Game Concepts To Books & Graphic Novels

By Tim B.

On 17 September 2024

In Other

Thought this was an interesting direction to explore with gen AI chatbots, first ChatGPT:

To adapt the open-world game concept to books or graphic novels, consider creating an interactive reading experience with the following elements:

Multiple Storylines: Readers can choose different paths that branch from a central narrative, much like side quests. Each path leads to a unique outcome or insight into the main story.

Non-linear Chapters: Similar to how players explore different areas of a game, readers can decide which chapters or segments to read first. These chapters can focus on different characters, locations, or events that tie into the overall plot.

Interactive Elements: Include prompts or optional tasks within the story, such as puzzles, decisions, or challenges, that impact how the plot develops or which character perspectives are explored.

Worldbuilding Details: Like in open-world games, the setting can be richly detailed. Include maps, lore pages, or sections that deepen the world, letting the reader explore background material at their own pace.

Re-readability: Design the book so multiple readings uncover new layers of the story. Different sequences of reading the chapters or tasks affect the reader’s understanding, creating a dynamic experience each time.

Graphic Novel Integration: Visuals can enhance the open-world feeling by depicting different paths, giving clues through illustrations, or showing how the story branches. Interactive panels may allow readers to ‘choose’ which scene or character arc to follow next.

The experience would feel immersive, allowing readers to engage with the content at their own pace, choosing the narrative depth and direction much like an open-world game.

See also: intextuality, hypertext, layered texts.

Claude gives something pretty similar for the most part, so we’re starting to see the thematic convergence between their training data sets on this I think:

That’s an interesting concept to explore. Let’s think about how we could adapt the open-world game format to books and graphic novels:

Non-linear narrative structure:

The book could have a central “main quest” storyline, but allow readers to explore side stories and subplots in any order they choose.

Chapters or sections could be designed to be read non-sequentially, with clear entry and exit points.

Modular storytelling:

The story could be divided into self-contained “quest” chapters or episodes that can be read in various orders.

Each module would contribute to the overall narrative but also work as a standalone story.

Choice-based progression:

At the end of each chapter or section, readers could be presented with multiple options for where to “go” next, similar to choose-your-own-adventure books.

These choices could lead to different side quests or back to the main storyline.

World-building focus:

Dedicate significant portions of the book to rich descriptions of the world, its inhabitants, and its lore.

Include “exploration” chapters that aren’t strictly plot-driven but allow readers to discover more about the setting.

Character interactions:

Introduce a large cast of characters that readers can “interact” with by choosing to read their specific storylines or backstories.

Visual mapping (especially for graphic novels):

Include detailed maps of the story world, allowing readers to visually track their progress and choose their next “destination.”

In graphic novels, use visual cues to indicate possible paths or points of interest.

Side quests and optional content:

Incorporate clearly marked optional chapters or sections that provide additional depth but aren’t necessary for the main plot.

These could be presented as separate short stories within the larger narrative.

Inventory and stat tracking:

Include a system for readers to track items, skills, or relationships acquired throughout the story.

This could be integrated into the narrative or presented as supplementary material.

Multiple endings:

Offer various conclusions based on the paths taken and choices made throughout the book.

Replayability:

Design the book so that multiple read-throughs yield different experiences and reveal new aspects of the story.

Would you like me to elaborate on any of these ideas or explore how they might be implemented in a specific genre or story concept?

Mistral Chat:

Translating the concept of open-world games to books and graphic novels involves creating a non-linear narrative structure that allows readers to explore the story universe at their own pace and in their preferred order. Here’s a format that combines elements from gamebooks, choose-your-own-adventure stories, and traditional novels/graphic novels:

Modular Narrative: Break down the story into self-contained chapters, sections, or issues that can be read in multiple orders. Each module could represent a specific location, quest, or character arc.

Branching Paths: Implement a branching narrative structure, similar to choose-your-own-adventure books. At the end of each module, provide multiple options for the reader to continue, allowing them to forge their own path through the story.

Side Quests and Optional Content: Include optional chapters or sections that contain standalone stories, side quests, or additional lore. Make these clearly distinguishable from the main story arc so readers can choose to engage with them at their leisure.

Non-linear Timeline: Present the story in a non-chronological order, allowing readers to piece together the timeline themselves as they read different modules.

Interconnected Storylines: Even though the story is non-linear, ensure that the various plot threads are interconnected. This encourages readers to explore different modules to uncover the full picture.

Tracking Progress: Include a simple system for readers to track their progress and the choices they’ve made. This could be as simple as a flowchart or a list of checkboxes at the beginning or end of the book.

Multiple Endings: Based on the reader’s choices throughout the story, provide multiple endings to increase replayability and encourage readers to explore different paths.

Worldbuilding and Lore: Enrich the open world with detailed descriptions, illustrations, and background information. This can be done through in-universe documents, maps, or supplementary materials.

For graphic novels, the visual element can greatly enhance the open-world feel. Here are a few additional ideas:

Interactive Maps: Include detailed maps that guide readers to different modules or sections based on their choices.

Visual Cues: Use distinct art styles or color palettes to differentiate between different locations, quests, or timelines.

Hidden Details: Encourage readers to scrutinize the artwork for hidden clues, secrets, or easter eggs that enhance their understanding of the world.

Examples of books and graphic novels that use some of these techniques include “Hopscotch” by Julio Cortázar, “S.” by J.J. Abrams and Doug Dorst, and “Watchmen” by Alan Moore and Dave Gibbons.

By incorporating these elements, you can create an immersive, open-world experience in a book or graphic novel format that engages readers in a unique and non-linear way.

A lot of these are obvious, some I don’t like that much, but it’s helpful to see them all laid out like this via 3 different sources.

Reflections on Threads Replies to Information Op-Ed

By Tim B.

On 16 September 2024

In Other

Currently at 1.3K angry comments over on Threads, which is a new record – even for me. I haven’t been following it closely, because you can’t really without an Instagram account, but so far, some notable hate coming in:

Michael Chabon emoji puked on my op-ed.

Elon Musk’s daughter wrote “delete this.”

Thanks everyone for taking time to comment!

I’m enjoying all the people who believe they are blocking me in responses. I guess the other Tim Bouchers on that site must be having a bad day. Sorry about that, eh?

Here’s an editorial cartoon I am struggling to make with Dalle, where a greedy author is eating books and crapping out money in a factory… It didn’t quite work as I wanted (after many tries) but good enough:

If you’re new here and are actually curious what this is all about, I recommend checking out my About page, and a recent Welcome New Readers post. That should keep you busy, and check out the thematic series in the nav bar at top.

If you’re just here to dunk on me, you’re still welcome. I made some memes for you to use. Knock yourself out.

Having worked for five years in content moderation and handling online complaints, I’ve heard it all before, and much much worse. And having had the audacity to merely use AI technologies for the last few years and be open about it as a creative person, I have *especially* heard it all before, and had it all directed at me.

I also noticed that there’s only one comment on the actual article’s post page online. I guess people who are paying for a premium subscription know the value of it, and aren’t going to spend their attention fighting about it.

Meanwhile, many/most of the Threads responders seem to not have read more than the headline and the preview text (if that) because it’s paywalled. Okay, totally understandable.

The most curious thing to me though is that it seems to be all quiet on Twitter as to this op-ed. This is the first time I’ve ever seen people flame me en masse from Threads, so that’s something I guess.

One person took the time to put together a post in the hopes of reaching my Google Alert, but don’t seem to have (found via Threads):

It’s because of people like you that the world of self published writing will be even worse off, and will be continued to be considered lesser than big house publishing.

I’ve written before about how to me the publishing industry is not worth defending. The fact that ‘big house publishing’ must even be our benchmark in the first places just proves its hegemony.

Ten years ago, people like East Bay Ray of the Dead Kennedy’s were railing against YouTube for engaging in what he called “sharecropping” tactics. These kinds of fights have been going on forever and will continue forever. Doesn’t mean we shouldn’t have them, it just means: the more things change…

If you want to critique AI, I am here for that. But let’s not stop there.

Anyway, here’s a free book download for all the new Threadsitors. It’s one of my favorites, The Politeness Protocols. There will be 10 free copies for the next couple days you can download if you want. You don’t even need to give your real email address.

“AI Guy” Meme Template

By Tim B.

On 15 September 2024

In Other

I thought it might be fun to “if you can’t beat ’em, join ’em” as to the whole situation with the enormous group of people mad at me for using AI on Threads.

I workshopped a meme concept in ChatGPT (4o) then took it into Dalle, which messed up the text. Then took it into Photoshop to generative fill over the bad text, then uploaded the finished underlying image template to Imgflip, where I hope people will be able to use it to take their potshots more effectively.

Here are some sample memes I made to get the ball rolling, one below and some more at the link (and a second set with more):

I know this represents what people think, and there’s no point in trying to persuade anyone differently, so… when in Rome!

And while you’re still angry, don’t forget to use this other set of custom memes I made in June to bash me. Cheers!

UPDATE: I made an alternate character with a second meme template here. Enjoy!

Is it too soon to start wearing this?

By Tim B.

On 14 September 2024

In Other

Because, normally the way it works in the “industry” is when a famous author emoji-pukes on you, that means you are now famous too, right?

The Critique of AI Doesn’t Go Far Enough

By Tim B.

On 13 September 2024

In Other

Another reply to the angry ~1K or so comments over on Threads…

If you want to critique AI, I am all for it. In fact, the entire basis of my AI Lore books is a massive critique of AI using the very same technology to show its flaws and foibles (not to mention its creative possibilities).

But an honest critique of the situation should not stop at AI training data. All of these people complaining to me about corporate exploitation of artists are doing so on platforms which are wholly corporate-owned, and which are in fact literally designed to harvest the free labor of anyone who agrees to take up the yoke of being a “content creator” for them. For all their time, energy, sweat, and tears, these platforms reward them with “likes and follows.” And then which, of course, use AI/ML to manage your recommendations, timelines, and predict and control your behavior on the platform, using your content however they want in a perpetual license. And let’s not forget about cell phones and the ridiculous inhuman patterns of addiction and exploitation that they enable…

But I don’t see anyone complaining about the extreme inhumane imbalance of any of that in comments, because they have all willingly adopted and accepted that form of corporate exploitation as their own. They have been blinded by the ubiquity of the hegemonic social media powers, and wrongly believe what they’re doing is somehow better and more noble than what I am doing, simply because it’s been around longer, and they believe that they can’t live without it.

But social platforms are founded on exploitation of creators, plain and simple. You might win some small benefit back from it financially in rare instances, but for most people all they win is a complex of anti-social personality disorders and an acute but vague swirling anxiety that never passes. I don’t blame people for acting inhumanely on platforms which create the conditions for them to do that, and encourage them with small, worthless rewards when they do.

But I do blame them for stopping the critique squarely at my front door. If you want to critique this system, go all the way. Strike the root. Don’t take a swing and then step back, pretending like the job is done. We’re only just getting warmed up…

Freedom of Expression Includes All Technologies (And People)

By Tim B.

On 13 September 2024

In Other

If you have a Threads/Instagram account, you can view this thread which has gotten totally out of hand on The Information’s link out to my op-ed on that platform. Or if you are like me, a conscientious objector to social media who believes it brings out the worst in people, you can at least see the last few vitriolic replies to my op-ed as a guest user, of which there are now 955 as of this counting, and which will undoubtedly be higher by the time I finish writing this post.

It occurs to me that ultimately this is a situation developing that touches strongly on freedom of expression, namely mine. The way the majority of people are reacting seems to be with the intent to shame or silence me – a position which I am now taking towards them, despite the mountains of verbal abuse they heap on me. They want me to not say what I am saying, feel what I’m feeling, or do what I’m doing. They want the things I explore, that I experiment with as an artist to fit within the narrow confines of their own definitions of what’s good and bad.

That’s their right to have and express those opinions. Just like it is mine to express this one. But I refuse to take up and carry the burden of the unhappiness of others. The simple fact that I am being open and above board about using new AI technologies does not suddenly nullify my right to free expression. The right of free of expression extends to all people, using all types of technologies to express themselves.

Discrediting Writers Who Work With Editors

By Tim B.

On 12 September 2024

In Other

Saw this comment over on Threads about The Information piece:

One is not an author if they don’t write every single word. That’s who we are. That’s what we do. If a computer is writing it for a person, said person is not the author. Or even a writer, really.

Said this yesterday, but I’ll say it again. Writing is often a highly collaborative process, especially when it’s done in a production environment.

For example, on the Information op-ed, I had strong feedback from two other readers, one of whom suggested language changes I ended up using. Then I had help from a press person to figure out who to pitch to, what my initial headline could be (it wasn’t used in the end), etc. Then I had help from an editor at The Information, who cut and slightly re-wrote parts of the piece, and gave me suggestions and asked questions that significantly improved the arguments and presentation. Then there was at least one copy-editor/proof-reader who came in to fix all those little buggy things. In other words, I did not write every single word.

So, in short, according to the person above’s criteria, I and every other single writer who has ever worked with editors, beta readers, proofreaders, etc. is “not a real writer.” That just doesn’t even make sense.

Meliorator & Brigadir: Mass Fake Account Management Software

By Tim B.

On 11 September 2024

In Other

This one slipped by my awareness, from July 2024, a PDF put out by the Joint Cybersecurity Advisory, authored by a bunch of different alphabet agencies. It describes a Russian state-sponsored software system to manage fake accounts en masse on social media platforms. The overall system is called Meliorator, and one of its components which I guess is the UI, is called Brigadir:

Brigadir serves as the primary end user interface of Meliorator and functions as the administrator panel. Brigadir serves as the graphical user interface for the Taras application and includes tabs for “souls,” false identities that would create the basis for the bots, and “thoughts,” which are the automated scenarios or actions that could be implemented on behalf of the bots, such as sharing content to social media in the future.

This is not the first time I’ve heard of systems like this. Did some pretty detailed work around this in a past life, visible in archived form here. Another more detailed 2017 long form research piece of mine was published here based on my looking into more of the actual tactics used by the Internet Research Agency. (I used to have that article hosted on my blog here, but I was seeing often reports from my hosting system that there were high numbers of Russian IPs attacking my site, until I took it down and they magically disappeared, mostly.)

That second linked article above tracked some quotes going back to a 2010 US Air Force for a solicitation for vendors to build a Persona Management System that has pretty much exactly the same product description as Russia’s Meliorator at its core, as described in PDF at top.

“Software will allow 10 personas per user, replete with background , history, supporting details, and cyber presences that are technically, culturally and geographacilly [sic] consistent. Individual applications will enable an operator to exercise a number of different online persons from the same workstation and without fear of being discovered by sophisticated adversaries. Personas must be able to appear to originate in nearly any part of the world and can interact through conventional online services and social media platforms. The service includes a user friendly application environment to maximize the user’s situational awareness by displaying real-time local information.”

Probably these kinds of ad hoc management systems have existed as long as people have been automating social media systems, which is presumably as long as they have existed. Now, of course, we get to throw AI into the mix and see what happens…

From a May 2024 article about OpenAI’s report of disrupting state actors using its fools for disinformation:

“All of these operations used AI to some degree, but none used it exclusively,” the report stated. “Instead, AI-generated material was just one of many types of content they posted, alongside more traditional formats, such as manually written texts, or memes copied from across the internet.”

Same old same old forever and ever.

I wonder when we’re allowed to look at these things from a more neutral lens than that of fixating on misinformation & disinformation, as bad as they can be. Like what if we started calling such endeavors “hyperreality” campaigns, and try to map them based on more complex sets of criteria? I’ve outlined something to that effect here. Narratologically, they make use of networked narratives and transmedia storytelling, and having a chance to see all this up close was very much at the beginning of how this art project of mine all got started. I’m interested in when these kinds of distributed storytelling systems can be open-sourced, and become simply another tool in a toolbox of communication and creative expression (aka “art”), instead of this use that is strictly bad or harmful. Maybe one day in a decidedly different form…

My thinking has always been, if everyone had a botnet, then the power of them would at least be widely distributed instead of concentrated in the hands of a few. People talk about teaching kids media literacy, but I never hear anyone saying we should teach them how to build botnets. Part of me wonders, if this future we’re heading towards might require them to have that kind of deep inside knowledge in order to counter other forces using those same techniques to push their own dominator hyperreality narratives. Just like they might need the skills and knowledge to be able to deter drone swarms in physical space.

Op-Ed in The Information

By Tim B.

On 11 September 2024

In Other

This opinion piece of mine came out in The Information today, and I’m really happy with the end result (paywalled, unfortunately – but trust me it’s good). Thanks to everyone who helped me on it, and all the pieces leading up to it to be able to communicate these ideas as effectively as I think was done here.

It made me think about how collaborative writing often ends up being with other people, and even if you have a lot of help on a piece from feedback or suggestions of others, and editors who might even re-write parts of your text, no one ever accuses you of being a “not real writer.” And yet, the process is in some ways the same as with working with AI tools. Except the problem is you don’t have the benefit of the judgement and lived experience of other humans helping to improve your communication. Which is not to say AIs don’t have a lot to offer – it’s just that’s the main thing they still lack.

If you’re coming in from reading that piece, I recommend checking out my About page for more. Many rabbit holes to follow there. Cheers!