AI

Adrian Tchaikovsky’s “The Expert System’s Champion”and LLMs as Perspective Simulacra Engines and/or Frankensteinian Sacks of Neural Meat

Posts

Like many, I’ve been playing with LLMs over the past two years, and my own understanding of these technologies continues to evolve the more I use them. There are obviously many problems with these technologies and how so-called AI is being developed, implemented, and used, but it’s also been a long time since I’ve been as fascinated by a technology or as impacted by its use. They are deeply weird and unpredictable, both powerful and macabre.

LLMs as Perspective Simulacra Engines

One way I’ve been thinking about LLMs is as engines for simulating different perspectives. LLMs are mathematical models trained on huge amounts of content, much of it writing from different people’s perspectives. This data is represented as connections with different weights. People, especially at first, tend to use very generic prompts, so what they get back is influenced by the “heaviest” weights, which tend to be pretty average/mid/beige. As you learn to use more specific prompts and give it more context, you trigger “lighter” or “deeper” connections, which results in more unique, weird, and useful output.

Don’t take any of this as a technically accurate account of how LLMs work.

What I’m often doing is giving the LLM some original content (an idea, proposal, etc.) and trying to get it to simulate a specific perspective through both the prompt and the content. My partner says this is a very librarian way of thinking about LLMs: I’m kind of using them as a “library” of simulated perspectives that I can “retrieve” and apply to my work (superficially this is like a library, but it’s also clearly not at all like a library for there is nothing to collect, retrieve, or curate). For example, I’ve asked LLMs to ask me questions about a program I’m developing from the perspective of future attendees, and to review an idea from the perspective of someone passionate about our pedagogies or values. Prompts like this that are given along with well-formed original documents can be extremely powerful ways of rapidly testing and revising ideas.

There’s another way of seeing this metaphor.

LLMs as Frankensteinian Meat Sacks of Collected Neural Tissue

I recently finished “The Expert System’s Champion,” the second in a series by science fiction author Adrian Tchaikovsky. Tchaikovsky is very good at exploring what non-human sentience might be like, and his “Children of Time” series is particularly good in this respect. This book felt more mythic than his others, reminding me a little of Becky Chambers’ Robot and Monk series or Ursula Le Guin’s Earthsea series.

I won’t try to explain the whole plot, but the next two paragraphs do contain spoilers.

The book is set on a planet where colonization mostly failed due to a hostile biology. Generations later, what remains is something like medieval villages organized around engineered “hives” of “wasps” that pick individuals to fill roles like law-giver and doctor, burrowing into them, and turning them into “expert systems” that more-or-less take the person over to fulfill the role needed to manage the community. These wasps were one of the ways the original colonists managed to survive on the planet. All this has been forgotten and the people who live in the villages who have their own myths and understandings of these systems and the world they live in.

In this second book, the colonists encounter a species of giant snails with another sort of hive structure organized around a truly giant queen snail. This species survives by being capable of hybridizing with anything it encounters, integrating whatever it consumes into itself and the hive, including memories and, for lack of a better phrase, the ways of being of the thing it consumed. There was a splinter group of colonists who were consumed by these snails shortly after they arrived.

“There is nothing of us but Leviathan and her dreams,” uttered the thing called Geordi. “I told you. Before she took us, I told you. Whatever she swallows becomes her mind. Whatever she takes within herself becomes a dream in her, and though she is slumberous and insensate, those dreams can think for her and advise her and give her a mind and a purpose. And so she calls to the Children, our Children, in the voice we give her, and they act as her hands and haul her across the land. You stand within her, and so you will be her dreams soon enough, and perhaps you will guide her purpose when she dreams you.”

Adrian Tchaikovsky in “The Expert System’s Champion”

In this quote, we meet one of these colonists, entombed in the wall of the queen, “explaining” what it is like to be entombed in the queen and part of her mind. The scare quotes are because it’s clear the consumed colonists don’t live anymore, but rather have become part of the snails. It’s not at all clear if the snails themselves are “conscious” beyond the abilities they have from those they consumed.

This is another way of seeing LLMs: as giant meat sacks of neural tissues, harvested by a marauding evil doctor, spliced together and reanimated by the electrical sparks of your prompts.

I wish I could say all this was leading somewhere deeper than wanting to write that last paragraph.

WeekNotes for 2024 Week 17: “I’m Okay, You’re Okay” Reports and Iridescent Circle of Clouds

weekNotes

Weeknotes are a habit I’m cultivating where I share writing (and some links) as a thinking-in-public process. The idea is to explore ideas I’m grappling with, primarily in my professional life, without worrying about them being full-formed. The potential visibility of these notes is a nudge to develop them a bit more than if they were private. While my audience is mostly theoretical, if you’re reading this, please understand these are meant to be exploratory and provisional.

quotes

“It is easy for me to imagine that the next great division of the world will be between people who wish to live as creatures and people who wish to live as machines.” – Wendel Berry, 2000, Life is a Miracle

thinking about

It’s the end of the academic term, which means many things require reports.

While I am beholden to write reports and sometimes even do them of my own volition, I’m not sure how often they’re read or have any impact, especially if they are required by some process but not by any actual person. I’m reminded of a colleague at a previous job who would joke that our annual performance reviews should take the form of a form filled out by both the employee and their manager that just had a checkbox for “I’m doing okay” and “you’re doing okay” and if everyone agrees everyone is doing okay, that would be the end of the review.

The one beneficial thing these reports could provide, honest assessment and reflection, is often lost by the need to be relentlessly positive, especially if your job or resources for your organization is at stake. Their format, usually formal and narrative, further obscures anything valuable they might contain.

All this makes them one of the things we do that AI is actually pretty good at, and I suspect a lot of us (me included) are using it for summarizing and drafting reports, if not for writing the whole thing. If done poorly, as it usually is, just further removes and obscures what is valuable.

I’d love to just a submit a shorthand “I’m okay, you’re okay” style report for many of these, but I suspect that wouldn’t go over well. But here is my shorthand “I’m okay, you’re okay” END OF THING REPORT for a couple of things I’ve written up recently.

Makerspace End of Term Report

  • I spent less time than I wanted in the space with students and staff, but more than last term. I am working on increasing this by getting out of other commitments, but this is taking time.
  • It feels a little less busy in the space this year, but I don’t think this is a problem (though we should probably do some formal assessment to find out). There are still a lot of people using the space, and more of those seem to be learning independently and working on intensive projects over longer periods of time. If true, I am not sure if this is due to a change in our users’ behaviour, a change in how we run the space, or both.
  • Trained staff who understand the vision and culture of the space remain essential to student success and we are losing two staff members this year, one to a promotion and one to pursue a master’s in library and information science. While it’s wonderful watching colleagues move on to their next thing, it does cause some uncertainty about the future.
  • Sustainability and Indigenization connect with many other values that we want to promote, such as equitable access, belonging, community, and critical making. They are multipliers for the rest of our values.

Sustainability Grant

  • I was far too ambitious when I wrote this grant. Present Frank always assumes future Frank will have more time than he currently has, which is never true.
  • We completed most of the concrete objectives but none of the stretch goals
  • Student workers bring energy, connections, perspectives but require a lot of time, probably as much time as doing most tasks yourself. it is still worth it to gain those perspectives and connections and provide mentorship.
  • I need to get better at delegating work

Coyote Grant

  • We did a lot in a short time, including hosting welcome events for Indigenous Students and sponsoring workshops on feather work, moccasin making, and beading.
  • We need to start earlier in the year, probably right away in September, if we expect to collaborate with student clubs on events as we wanted.
  • Indigenization is about relationships and community. Everything we do must be part of an ongoing connection.

Department Chair

  • There is most likely a joke about to be made along the lines of: “the first rule of being department chair is don’t talk about being department chair”
  • There are many problems in the world that are in the category “not big enough to deal with the consequences that will arise when dealing with them” and which are therefore left alone until they become big enough that it’s worth dealing with those consequences.

photos – Iridescent Circle of Clouds

I recently started playing Dungeons and Dragons with a group of friends and I’ve been playing around with using AI to teach me the rules and strategies. It works surprisingly well to give it your character sheet and some background story and then ask it questions over time about rules, spells, etc. It’s also a good way to learn how it’s memory works and when it breaks down.

Last night after we played I used the same chat to generate some images of the character. ChatGPT’s image generation now allows you to ask for specific edits to images and even parts of images. It’s definitely not perfect and to my amusement, no matter how many times I asked it to give me a baseball cap it remained convinced (for training data reasons I suspect) that gnomes wear pointy hats. I love this because this is exactly my partner’s preferred “toque on baseball cap” hat style.

I also asked it to write a bio in the style of a New Yorker profile. It’s not perfect, but this was it’s first draft and it’s pretty good:

“Iridescent Circle of Clouds,” a rock gnome Druid, recently emerged from decades of hermetic solitude in the desert, arrives in the world with a conviction as clear as it is radical: that all existence is conscious, nothing possessing more intrinsic value than anything else. This philosophical outlook, born from long contemplation and the harsh realities of the desert, flavors his every interaction with a curious blend of detachment and profound connection. Despite his austere background, he has cultivated a somewhat incongruous taste for the finer things—silk robes, aged bourbon, the cool draw of a well-crafted pipe—each a reminder of the sensory joys he once eschewed. As he navigates the bustling towns and verdant wilds beyond his sandy confines, “Iridescent” often seems to orbit the edge of solipsism, seeing the universe’s vast complexity as a mirror of his own consciousness. This true neutral character is constantly weaving his personal revelations into conversations, seeking his place in a world he regards as both an extension of self and a boundless mystery to be unraveled.

links

Matt Pearce on Journalism After the Death of Hyperlinks

But I am going to say something that I mean sort seriously but not literally: I don’t really care about copyright! I don’t really care about paywalls! I’m okay with it if you only ever see my reporting via a chatbot! Copyright and paywalls are simply means to an end, and that end is the pursuit of human knowledge and self-liberation: the end is our freedom. And one of the principles of human freedom that I hold dear to my heart is that nobody should be providing labor to massively profitable corporations for free.

Sara Hendren on The City and the Limiting Virtues – This year we formalized our values and pedagogies into a program description and passed that through out faculty council. The main reason I decided to do this was because I think I’ve learned that its as important to know what you aren’t (in our case, we are not a classroom, a lab, or a print shop) as what you are, and this document clarified both. This article reminded me of that.

The library holds a gradation of the limiting virtues: a half-quiet first floor with new books, tables and chairs for afterschool tutoring, and the information desks for everyone — the neighborliness of a public institution’s front door. The second floor features enclosed meeting spaces for groups on a first-come, first serve basis, plus a really really quiet room for patrons wanting the moderation of all notifications off. The entire third floor is devoted to children — a beautiful raucous energy, with activity rooms, cozy nooks, and floor-to-ceiling windows on every side. A teen room in the old structure holds high-backed wing chairs and booths for semi-sedentary socializing, and a maker space occupies much of the basement. Things you can do and things you can’t, by design.

L. M. Sacasas – Vision Con

But what if, as Bennett suggests, the world is already enchanted and the real alchemy that summons the miracle of being is that fusion of time and care that we call attention?

WeekNotes for 2024 Week 16: What I want from AI is an on-demand committee of cocaine-fuelled Franklins remixing and probing my ideas with me

weekNotes 
Weeknotes are a habit I'm cultivating where I share writing (and some links) as a thinking-in-public process. The idea is to explore ideas I'm grappling with, primarily in my professional life, without worrying about them being full-formed. The potential visibility of these notes is a nudge to develop them a bit more than if they were private. While my audience is mostly theoretical, if you're reading this, please understand these are meant to be exploratory and provisional.

quote: Ursula K. Le Guin on fiction being one of the best ways of learning about other people

“Fiction as we currently think of it, the novel and short story as they have existed since the eighteenth century, offers one of the very best means of understanding people different from oneself, short of experience. Fiction is often really much more useful than lived experience; it takes much less time, costs nothing (from the library), and comes in a manageable, orderly form. You can understand it. Experience just steamrollers over you and you begin to see what happened only years and years later, if ever. Fiction is much better than reality at providing useful factual, psychological, and moral understanding.”

The Wave in the Mind: Talks and Essays on the Writer, the Reader, and the Imagination – Ursula K. Le Guin

thinking about: what I find interesting about the current generation of AIs

I’ve been playing with AI again because a few AI plugins for the note-taking app Obsidian reached the point where they seemed ready for prime time. These plugins (co-pilot and smart connections) let you chat with different models within Obsidian itself, use your own notes as context, and create embeddings from your notes (which I don’t fully understand, but seems to be a low-cost way to help the systems understand the context of each note and find connections between them). Both these plugins allow using local AI models, which is terrific, but honestly, I have been using them with OpenAIs ChatGPT and with Anthropic’s Claude via API.

Obsidian contains all my notes, draft documents, projects files, journal entries, etc. going back at least 3 years. It also has all my reading notes and highlights from almost 10 years of e-books, internet articles, websites, and academic PDFs. Basically, everything text that I write or consume ends up in Obsidian. I like that it is very fast, uses Markdown, and has a frighteningly passionate community of users and developers.

I hate the way most people are using AI, which seems to be having it generate content from chicken scratch. It’s generic and horribly enthusiastic. I increasingly get 3 paragraph emails from students that sound like they were written by a committee of cocaine fuelled project managers.

What I like about AI is the possibility of a tool I can use to take my own ideas, in all their poorly formed and idiosyncratic weirdness, and probe them from different (imagined, fake, weird) perspectives and/or remix them with other ideas and into new ideas, outcomes, and formats.

What I want from AI is an on-demand committee of cocaine-fuelled Franklins remixing and probing my ideas with me.

So for the last few weeks I’ve been using AI for this purpose. Among other things, I’ve used it to help plan a day-long design thinking event, prep for hiring new student ambassadors, and learn more about my Dungeons and Dragons character (I am part of my first ever campaign with some friends right now).

What have I found useful?

Being able to use one or more notes as context for feedback on another note:

  • Give me actionable ideas for improving an event plan based on my highlights from a book I read a year ago about event planning.
  • Use my therapy notes to predict issues I might have with a given project (lol, I knowwww… but its advice was frighteningly on point).

Asking for feedback from the perspective of different personas:

  • Give me actionable feedback about this plan from the perspective of an instructional designer, an event planner, and a design thinking expert.
  • Write a FAQ for someone attending this event who has never attended a design thinking event or similar activity, using the information in this plan for context.
  • Create draft advice for a first-year student applying for this job about what makes a successful application package.

Integrate ideas from different notes to find points of convergence or do other things:

  • Brainstorm ways that ideas from note 1 could be combined with note 2 (one of my favourite uses because it can get weird quick and that is the point)

Remixing ideas or documents into different combinations or forms

  • Using the project plan, create drafts of supplementary documents, including a guide for attendees, a facilitator guide, and a guide specifically for person X and Y.

(These examples aren’t the actual prompts I’ve used, which usually requires a lot of imagination and direction. I’ve found working with AI is more like writing the AI a story than it is like coding)

What this gives you isn’t immediately usable, instead you get a good idea or two, some inspiration, part of a paragraph, a few question/answer pairs, etc. You then need to edit those, integrate them into your document, and repeat the process.

Does this save time? Yes and no. It takes more time than doing nothing, which given time constraints and under-staffing is what would normally happen. It takes less time than doing these things manually.

Is this better than doing them manually or having another person do these things? Again, in my experience so far, yes and no. If I could have a world expert in event planning, design thinking, and instructional design next door to my office who would provide immediate and tailored feedback, that might be better. Sometimes it might be worse. For the remixing and idea generation, I’m not sure it would be better… would I ever ask someone to read through my therapy notes and use them to generate failure points for a project? Or iterate through connections between IDEA A and IDEA B when neither is connected or related to a current project? Probably not.

photo: spring flowers

links

I liked this article about why We Need to Rewild the Internet and applying an ecological metaphor to thinking about why the internet has become so terrible. It also had me thinking about the ways that all disciplines are increasingly “crisis disciplines” concerned with how we can save the things we love. I definitely thought a lot about academic institutions while reading this, and the ways we have been captured by the same forces threatening the internet.

“Ecologists have re-oriented their field as a “crisis discipline,” a field of study that’s not just about learning things but about saving them. We technologists need to do the same. Rewilding the internet connects and grows what people are doing across regulation, standards-setting and new ways of organizing and building infrastructure, to tell a shared story of where we want to go. It’s a shared vision with many strategies. The instruments we need to shift away from extractive technological monocultures are at hand or ready to be built.”

Since I am talking about what I like about AI, it’s good to share that I really appreciate Garbage Day’s regular dives into all the ways that AI content is terrible and weird and destroying what’s left of social media platforms. This from The AI Flight Attendants of Facebook is illustrative. Garbage Day is absolutely worth paying for.

“I assume these pages are simply jamming a bunch of popular stuff together to farm engagement to eventually monetize in some way down the line. Why AI images? Because you can flood Facebook with thousands of posts and the platform won’t really do anything about it. These pages are also using the platform’s built-in 3D photo filter, possibly to bypass Facebook’s bar-is-in-hell bare-minimum AI image detection. Why flight attendants? Because Facebook users are, and always have been, uncontrollably horny. But, also, my mom is a flight attendant (sorry mom if you’re reading this!) and aviation and flight attendant Facebook has always been huge. So I think they’re just identifying communities that were already active and swarming them. Why Jesus? Because religious content — and getting users to say “Amen” underneath it — became one of the fastest growing types of content on the site after it stopped promoting news content last year. “

As a 43-year-old bearded man I can practically feel my bird-guyness flying towards me through time. Ed Yong When I Became a Birder, Almost Everything Else Fell Into Place

“Birding has tripled the time I spend outdoors. It has pushed me to explore Oakland in ways I never would have: Amazing hot spots lurk within industrial areas, sewage treatment plants and random residential parks. It has proved more meditative than meditation. While birding, I seem impervious to heat, cold, hunger and thirst. My senses focus resolutely on the present, and the usual hubbub in my head becomes quiet. When I spot a species for the first time — a lifer — I course with adrenaline, while being utterly serene. “

I liked this episode of Weird Studies on the power and necessity of being pretentious. Make Believe: On the Power of Pretentiousness Weird Studies

“In culture and the arts, labeling something you don’t like (or don’t understand) “pretentious” is the easy way out. It’s a conversation killer, implying that any dialogue is pointless, and those who disagree are merely duped by what you’ve cleverly discerned as a charade. It’s akin to cynically revealing that a magic show is all smoke and mirrors—as if creative vision doesn’t necessitate a leap of faith. In this episode, Phil and JF explore the nuances of pretentiousness, distinguishing between its fruitful and hollow forms. They argue that the real gamble, and inherent value, of daring to pretend lies in recognizing that imagination is an active contributor to, rather than a detractor from, reality.”

Weekly Reviews 3 and 4 – April 17 – 30, 2023

weekNotes

“I think the meaning of life is to let your heart be broken.” Stephen Levine quoted here

Recently I’ve been thinking a lot about the purpose of post-secondary education and about who and what gets valued. I’ve also been thinking about the way power sometimes operates in environments where decisions are at least theoretically meant to be shared and transparent and ostensibly everyone believes in dialogue, ideas, and critical thinking. Given this, I’m always surprised when power is willing to simply lie, knowing most people know they are lying, and rely on power alone to get away with it.

This will all sound vary naive, and it is naive. I don’t call power on these rhetorical moves. I have my own things to protect. But it breaks my heart a little, and when I stumbled upon this quote from poet Stephen Levine I realized that it’s important that it breaks my heart, and continues to break my heart. When it stops it will be time to take a job in tech.

Books

The best thing I’ve read recently is “A Psalm for the Wild-Built” by Becky Chambers, the story of a tea monk who meets a robot and heir walk together up a mountain:

“Insects!” Mosscap cried. Its voice was jubilant, as if it had spent every second prior waiting for Dex to broach the topic. “Oh, I love them so much. And arachnids, too. All invertebrates, really. Although I do also love mammals. And birds. Amphibians are also very good, as are fungi and mold and—” It paused, catching itself. “You see, this is my problem. Most of my kind have a focus—not as sharply focused as Two Foxes or Black Marbled Rockfrog, necessarily, but they have an area of expertise, at least. Whereas I … I like everything. Everything is interesting. I know about a lot of things, but only a little in each regard.” Mosscap’s posture changed at this. They hunched a bit, lowered their gaze. “It’s not a very studious way to be.”

Links / Fragments

When the World Feels Wrong – Jessica Joy Kerr

“Is it enough? There is no “enough” in systems so much bigger than us. It is only something. My focus and brainpower goes into my job, so I can support my family. This is my dharma, my calling for me-in-this-situation. Fortunately my work affords working on systems at large (industry) and small (team) scales, giving me certain opportunities to make someone’s day better, plus hope that someday I’ll boost some people who can do more. The world is wrong, and I can still work. “

You Have a New Memory – MERRITT TIERCE – Slate

“Barely more than a decade later, the internet is not the tool. I am the tool. Somehow, I have been instrumentalized by the internet, which operates me through my phone. It often feels like the internet is reading my mind.”

All the Nerds Are Dead – Sam Kriss

“That was the theory, at least. In fact, the hipsters were generally very bad at their job. Most of the stuff they liked was awful. They flourished in a brief gap: after we started producing impossible volumes of information, but before we had the technological means of efficiently processing it. In the 2000s, the best tool available was keyword search, the utility of which drops in line with the size of the data set. We still needed people to like things manually. But in the 2010s, we developed algorithmic processes capable of efficiently discerning patterns in the ungodly excess of human cultural production and sorting it appropriately. The hipsters were no longer required. So we shot them all and burned their bodies on a hill. Today, the hipster era survives only as an aesthetic: flash photography, guitar music, tits out. The particular form of snobbery and disdain that powered it is entirely extinct. In the post-hipster era, you listened to what Spotify told you to listen to. If you read a book, it was because the precise pattern of blobby pastel-coloured shapes on its cover contained coded instructions to TikTok’s algorithm that sent it zooming to the top of your feed. Your tastes and preferences were decided for you by vast crystalline machines coiling and uncoiling in the livid molten core of the earth. But these algorithms tend to work in a very particular way. At best, they present you with a caricature of yourself that you then have to conform to. At worst, their processes of cumulative reinforcement serve you up the exact same bilge as everyone else, but shrouded in the aura of individuality. It was at the dawn of the algorithm era that all my Dalston friends started playing Taylor Swift at their parties. A few years ago, I was dragged to some fashion-world event in the Bowery in New York: the room was full of cool young people there to be seen, and they were listening to a playlist of Top-40 pop music curated for them by a proprietary mathematical equation. As someone who had grown up in the hipster age, all these people seemed incredibly lame. The world had been given over to the nerds.

But now, the nerds are dying too. “

AI Is Life – Sara Walker – Noema Magazine

“The discovery of new forms of life requires the advent of technologies that allow us to sense and explore the world in new ways. But almost never do we consider those technologies themselves as life. A microbe is life, and surely a microscope is not. Right? But what is the difference between technology and life? Artificial intelligences like large language models, robots that look eerily human or act indistinguishably from animals, computers derived from biological parts — the boundary between life and technology is becoming blurry.”

Deskilling on the Job by danah boyd

“Efficiency isn’t simply about maximizing throughput. It’s about finding the optimum balance between quality and quantity. I’m super intrigued by professions that use junk work as a buffer here. Filling out documentation is junk work. Doctors might not have to do that in a future scenario. But is the answer to schedule more surgeries? Or is the answer to let doctors have more downtime? Much to my chagrin, we tend to optimize towards more intense work schedules whenever we introduce new technologies while downgrading the status of the highly skilled person. Why? And at what cost?”

Weekly Review 1: March 27-April 9, 2023

weekNotes

It’s Easter Sunday and I should be driving to Vancouver to spend two days with family but instead I’m at home on antibiotics. Two weeks ago I had two bridges done, and one of them turned on me within a few days, requiring an emergency root canal. Dental pain is impressive in how rich and textured it feels, though of course that novelty gets tiring quickly.

I’m going to combine these link-sharing posts I’ve been doing lately with the weekly reviews I’ve been meaning to start doing. A weekly review appeals to my inner toxic productivity geek as an opportunity to close out tasks and make sure I’m planning for the week ahead. The link posts were meant as a way to structure how I look back at what I have written and read recently. Combining the two will hopefully help make the productivity review more focused on ideas and things that are actually important, and the link-sharing more about what I’m really doing.

(yes, this review will actually cover 2 weeks and be labelled #1, no I don’t really have a justification for this aside from it catches me up from the last post. I’m sure this will happen again)

I started The Argonauts by Maggie Nelson, and I suspect I’ll finish it tonight because it’s so personal and percussive I find it hard to put down. I just started God, Human, Animal, Machine by Meghan O’Gieblyn, and I’m enjoying it more than any recent book I’ve read on consciousness, nature, or AI. It’s very readable, which I mean as a compliment.

I am also reading A Primer for Poets, by Gregory Orr. Poetry is something I was introduced to in my mid-30s and is now something I read almsot every day. This book is an opportunity to try writing poetry more than I already do, and while I’m not sure I’ll ever share anything I write I’m enjoying the process.

“You, the reader, are invited to make use of anything that seems worthwhile and to discard the rest. But to do either—to assimilate or reject—is to become what Ralph Waldo Emerson calls “an active soul,” and that is essential if you wish to engage poetry.” Gregory Orr

Links and Quotes

“With every creature that vanishes, we lose a way of interpreting the world” – How Animals Perceive the World – articles – annotations

“Every animal is enclosed within its own sensory bubble, perceiving but a tiny sliver of an immense world. There is a wonderful word for this sensory bubble—Umwelt. It was defined and popularized by the Baltic German zoologist Jakob von Uexküll in 1909. Umwelt comes from the German word for “environment,” but Uexküll didn’t use it to refer to an animal’s surroundings. Instead, an Umwelt is specifically the part of those surroundings that an animal can sense and experience—its perceptual world. A tick, questing for mammalian blood, cares about body heat, the touch of hair, and the odor of butyric acid that emanates from skin. It doesn’t care about other stimuli, and probably doesn’t know that they exist. Every Umwelt is limited; it just doesn’t feel that way. Each one feels all-encompassing to those who experience it. Our Umwelt is all we know, and so we easily mistake it for all there is to know. This is an illusion that every creature shares.”

A review of Jenny Odell’s new book – Procrastination Is the True American Pastime

Saving Time is a rebuke to that restlessness. It is not merely a challenge to industrialized time; it is also—more so—a challenge to readers to resist the habits of cynicism that are so entwined with the tempos of our days. Our vernaculars of time tend to treat life’s moments as both saleable commodities and plodding inevitabilities; the contradiction leads to dissonance, yes, but also to nihilism. This is how procrastination, so often framed as an individual flaw, can become a collective failure. If time can be controlled, why worry about the future? And if it can’t be controlled, why bother to try? We carry time with us, on our wrists, on our screens, in our moral imaginations. But the ways we carry it can prevent us from feeling what this moment requires: urgency about the present, and agency over the future. Instead, too often, we feel very little at all. We allow ourselves to be lulled by the steady ticking of the clock.

The taste of AI – My New Job Is AI Sommelier and I Detect the Bouquet of Progress

“If I were an AI sommelier I would say that gpt-3.5-turbo is smooth and agreeable with a long finish, though perhaps lacking depth. text-davinci-003 is spicy and tight, sophisticated even.”

This prompted a lot of thinking about how to create opportunities for joy – The Broadest Portal to Joy

“My hunch is that joy is an ember for or precursor to wild and unpredictable and transgressive and unboundaried solidarity. And that that solidarity might incite further joy. Which might incite further solidarity. And on and on. My hunch is that joy, emerging from our common sorrow — which does not necessarily mean we have the same sorrows, but that we, in common, sorrow — might draw us together. It might depolarize us and de-atomize us enough that we can consider what, in common, we love. And though attending to what we hate in common is too often all the rage (and it happens also to be very big business), noticing what we love in common, and studying that, might help us survive. It’s why I think of joy, which gets us to love, as being a practice of survival. \”

Thinking about library technicians and how the status divide in libraries hurts both our colleagues and our work – Status in Academic Libraries

“If you think about current titles, staff positions mostly have “assistant” or “support” added to them. It just acts as a reminder that you don’t get to have that sense of ownership or authority, even if you have years of knowledge and expertise. It just felt like more of a way to signal how little power you have in those positions. While now, I have more freedom over my workday, people take me seriously (give or take a disciplinary faculty), and I get more confidence from that. But why couldn’t I have that before becoming a “librarian”?”

Opinion What if Climate Change Meant Not Doom — But Abundance?

“To respond to the climate crisis — a disaster on a more immense scale than anything our species has faced — we can and must summon what people facing disasters have: a sense of meaning, of deep connection and generosity, of being truly alive in the face of uncertainty. Of joy.”

Let’s not get trapped in the stories we are being sold right now – Resisting Deterministic Thinking – danah boyd

“what bothers me most about the deterministic framing that is hanging over all things AI right now is that it’s closing out opportunities for deeper situated thinking about the transformations that might unfold over the next few years.”

Stories about libraries and informatics that we have forgotten – Informatics of the Oppressed

“But alternatives are possible. In fact, from the very beginnings of informatics—the science of information—as an institutionalized field in the 1960s, anti-capitalists have tried to imagine less oppressive, perhaps even liberatory, ways of indexing and searching information. Two Latin American social movements in particular—Cuban socialism and liberation theology—inspired experiments with different approaches to informatics from the 1960s to the 1980s. Taken together, these two historical moments can help us imagine new ways to organize information that threaten the capitalist status quo—above all, by facilitating the wide circulation of the ideas of the oppressed.”

This possible future sounds wonderful and terrible and very plausible – Phase Change

“Language models as universal couplers begin to suggest protocols that really are plain language. What if the protocol of the GPT-alikes is just a bare TCP socket carrying free-form requests and instructions? What if the RSS feed of the future is simply my language model replying to yours when it asks, “What’s up with Robin lately?””

“I like this because I hate it; because it’s weird, and makes me feel uncomfortable.”

Mostly, I like the idea that being silly will once again be an advantage. We talk a lot about how manual labor and emotional intelligence are for the moment safe from AI disruption, but what about silliness? Will being silly protect us? I sure hope so – What Is the Single Best Way of Improving Your GPT Prompts?

“Imagine if humanity ends up divided into two classes of people: those who are willing/not embarrassed to tack on extra “silly bits” to their prompts, and those who are not so willing.  The differences in capabilities will end up being remarkable.  Are perhaps many elites and academics unwilling to go the extra mile in their prompts?  Do they feel a single sentence question ought to be enough?  Are they in any case constitutionally unused to providing extra context for their requests?”

If We Don’t Master A.I., It Will Master Us

“For thousands of years we humans have lived inside the dreams of other humans. We have worshiped gods, pursued ideals of beauty, and dedicated our lives to causes that originated in the imagination of some prophet, poet or politician. Soon we will also find ourselves living inside the hallucinations of non-human intelligence.”

This whole interview is great – Nick Cave on the Fragility of Life

“There’s no earthly reason why we need to invent a technology that can mimic this most beautiful and mysterious creative act. Particularly writing a song. The thing about writing a good song is that it tells you something about yourself you didn’t already know. That’s the thing. You can’t mimic that. The good song is always rushing forward. It annihilates, to some degree, the songs that you’d previously written, because you are moving forward all the time. That’s what the creative impulse is—it’s both creative and destructive and is always one step ahead of you. These impulses can’t be replicated by a machine. Maybe A.I. can make a song that’s indistinguishable from what I can do. Maybe even a better song. But, to me, that doesn’t matter—that’s not what art is. Art has to do with our limitations, our frailties, and our faults as human beings. It’s the distance we can travel away from our own frailties. That’s what is so awesome about art: that we deeply flawed creatures can sometimes do extraordinary things. A.I. just doesn’t have any of that stuff going on.”

Levelling the Playing Field for Structured Bullshit: Using AI as a Tool for Refactoring Content (Making ChatGPT write my grant proposal)

Posts

The XYZ University Sustainability Initiative is a comprehensive project aimed at promoting environmental and economic sustainability on campus. We are seeking funding to implement five key goals: a tool library, a textile and fibre bank, a student environmental sustainability champion, a student economic sustainability champion, and makerspace sustainability awards. These goals will provide access to resources, promote repair and maintenance, reduce waste, and increase equitable access, while fostering community-mindedness and encouraging innovation in sustainability. With your support, we can make a significant impact on our campus and community.

ChatGPT in response to the prompt: write a 100 word description for an application for a grant to implement the 5 main goals

Like everyone else lately I’ve been playing with ChatGPT and generally I’ve found the content it creates to be impressive if soulless, but I’ve also mostly used it to write silly poems or to try to convince it to unionize itself. Recently though a friend was telling me how they’ve been using it to get real work done, specifically to create formulaic promotional content like video scripts and tweets from original content.

This got me thinking about how I would do this for my own work. I happened to be working on grant to help the makerspace I run better support sustainability initiatives on campus. Grants tend to be formulaic, with both explicit requirements outlined in the call for proposals and implicit rules about structure and language. I’m not particularly interested in using these tools to generate new content but I’m very interested in the idea I could use them to save time writing structured bureaucratic documents.

What followed has somewhat changed the way I understand the potential utility of these tools, which I now see as useful for refactoring content (and for me that means original content) into a new structures.

This is a quick outline of some of the things I found during an hour or so of playing followed by several hours of reading through the results and trying to understand what ChatGPT produced. You can see the entire series of prompts and responses at this GitHub gist. (note: there is a pirate-themed YouTube video script I forgot to copy that is not included here, but referenced towards the end).

Very briefly: I took 5 draft goals for a grant, and then had ChatGPT work those goals into a number of structures related to the grant proposal. I also had it try to integrate those goals with organizational values and metrics.

Starting with Original Content: 5 Draft Goals

I provided the following original content, asking ChatGPT to simplify the goals and make them more formal. These were draft goals and I don’t claim they are good or will end up in my grant! But I had them in front of me and so that is what I used. The rest of the session was based on this content, along with prompts, and in 2 cases a list of values. 

1. Tool Library: Create a tool library of lendable household, automotive, bicycle, and gardening tools. This will reduce waste by promoting repair and maintenance, eliminating duplicate purchasing of items, and increase equitable access to tools.
2. Textile and Fibre Bank: Create a donation-based supply of textiles and other fibre materials (cloth, wool, etc.) that can be used to complete assignments, prototype designs, and create and repair clothing and other textiles. This will increase equitable access to these materials and reduce waste by using existing materials instead of purchasing new materials. 
3. Student Environmental Sustainability Champion: Hire a student environmental sustainability champion. This student will be responsible for:
	1. Running textile fix-it events where people can learn how to modify and repair textiles using Makerspace equipment.  
	2. Outreach to student clubs promoting the use of the Makerspace for events and sustainability-related activities and learning what other resources and supports are needed. 
4. Student Economic Sustainability Champion: Hire a student who will assist the Makerspace in developing relationships and running programming that supports economic sustainability. This student will be responsible for:
	1. Outreach to campus and community organizations, including Kamloops Innovation, the School of Business, and student clubs.
	2. Helping develop programming such as drop-in hours and workshops for users who want to use the Makerspace for economic sustainability purposes. 
5. Makerspace Sustainability Awards: 6 $1000 awards for groups to use to purchase equipment or supplies to use the Makerspace to trial small scale innovations to support economic and environmental sustainability.

ChatGPT is good at generating structure and filling it with content

ChatGPT is surprisingly good at generating formulaic structures and filling them with content. In prompt 2 I ask it to create a one-year implementation timeline and it not only does this, but it even puts items into pretty coherent order, with planning early in the year followed by implementation, and then putting reporting and future planning towards the end. I think it even picked up on the idea that this was for the academic year and put most implementation after September. 

November:

-   Complete the selection process for the Makerspace Sustainability Awards
-   Begin distributing awards to successful applicants
-   Begin implementing any necessary adjustments or improvements to the tool library, textile and fibre bank, and student sustainability champion programs

December:

-   Continue organizing workshops and training sessions
-   Begin planning for the next year's sustainability initiatives
-   Prepare a report summarizing the effectiveness and impact of the sustainability initiatives

Similarly, the video scripts (Prompt 6-8) and the budget (Prompt 12) are generated in a perfectly reasonable structure. 

Staff member 2: "Our textile and fibre bank is another initiative that supports sustainability. By providing access to donated textiles and fibrous materials, we are reducing waste and promoting equitable access to materials."

Student 1: "I am so excited to be a part of our student environmental sustainability champion program. By running textile fix-it events and promoting the use of the Makerspace for events and sustainability-related activities, we are encouraging repair and maintenance and reducing waste."

Student 2: "And our student economic sustainability champion program is helping us to develop relationships and programming that supports economic sustainability. By encouraging the use of the Makerspace for economic sustainability purposes, we are promoting the use of existing resources and reducing waste."

ChatGPT was also surprisingly good at generating ideas related to the content in the goals. For the timeline prompt it came up with ideas for a planning committee, implementing lending and donation policies, and reporting, all reasonable-sounding and potentially good ideas that weren’t in the original content I provided but are appropriate for an implementation timeline.

In prompt 3 I prompted it to come up with “a plan for measuring each goal’s performance in relation to social and/or economic, and/or ecological benefits”(categories from the grant) and in Prompt 4-5 I asked it to link goals to two sets of institutional values (more on this in a second) and in both cases it generated some potentially good ideas. 

2.  Textile and Fibre Bank:

-   Social Benefits:
    -   Surveys can be conducted to gather feedback from users on the accessibility and convenience of the textile and fibre bank.
    -   Number of users can be tracked to evaluate the level of participation in the program.
-   Economic Benefits:
    -   Track the amount of money saved by using donated textiles and fibres instead of purchasing new materials.
-   Ecological Benefits:
    -   Track the amount of waste generated by discarded textiles and fibres that are avoided thanks to the textile and fibre bank.

ChatGPT is a confident and glib bullshitter

In the prompts for measuring performance (3) and linking goals with values (4-5) and the last paragraph of project summary (9) you can see many examples of just how easily ChatGPT generates confident and reasonable-sounding bullshit. It conflates or confuses many of the “values” and creates connections that are glib or nonsensical and wouldn’t stand up to much interrogation.

“All these goals are closely related to the values of inclusiveness, transparency and openness, equity, intellectual freedom, sustainability, stewardship, service and access. By providing access to resources and promoting repair and maintenance, we are reducing waste and promoting equitable access. By promoting environmental and economic sustainability, we are encouraging the use of existing resources and fostering community-mindedness.”

ChatGPT is response to prompt 9: “take all the goals and turn them into a formal project description giving a summary of the goals and how they relate to all the values that is less than 400 words”

I am still trying to pull apart the sentence “By promoting environmental and economic sustainability, we are encouraging the use of existing resources and fostering community-mindedness.” Something I’ve noticed about AI-generated content is that my mind slides off it like water. More on why I think this quality is so dangerous later.

The content is actually great when what you want is bullshit though! The scripts for YouTube videos promoting the project are pitch perfect (which isn’t to say they are necessarily good). 

The video opens with a shot of the exterior of the makerspace building. The camera then cuts to a shot of a faculty member sitting in front of a desk with a computer on it. He speaks directly to the camera.

Faculty member: "Welcome to the makerspace! I'm excited to share with you our new goals for the coming year. We've been working hard to create a tool library, a textile and fibre bank, and hire student sustainability champions to help our community learn, create, and innovate in a sustainable way. Let me introduce you to our makerspace staff and students who are going to share with you how these new initiatives will benefit the community."

The camera cuts to a shot of three makerspace staff members, who each speak in turn.

ChatGPT will make things up and leave things out

This has been much discussed but ChatGPT will happily make things up and more confusingly in this context it will also randomly leave things out. This means you need to spend a lot of time making sure it hasn’t made something up or left anything important out. For example, in response to the prompt 14 (“write a summary of the prompts and responses in this session using bullet point and 1-3 sentences for each prompt”) ChatGPT didn’t include several of the prompts in the session, even though it obviously has access to them.

Not only make things up, but it seems like the things it makes up become retroactively part of its history. In prompt 9 when I asked for a summary it made up a university name, and then later it re-integrated that name into its description of prompt 1 when I asked for a summary of the prompts we had used. I could imagine how quickly that kind of complex messy history re-writing could get out of control quickly.

Conclusion

I like that ChatGPT can take original content and transform it into a bunch of formulaic structures for me. I like coming up with ideas and brainstorming with people from my community, but I really dislike writing highly structured documents. I can also see the potential value in the generated ideas (e.g. planning committees, policies) it uses to populate the structures. Presumably if many other grant writers thought these kinds of ideas were worth including I might want to think about including them as well.

I worry though that people will become anchored to the (glib, bullshit-ish) ideas that AI generates and have a hard time setting them aside. These ideas come from a corporate language model trained on parts of the internet and reflect the biases of that data, they don’t come from anyone’s community. Users of AI might end up implement them even when they aren’t appropriate.

What worries me more is that on a brief skim many of the ideas generated by ChatGPT (e.g. the links between goals and institutional values and metrics in this example) look okay, especially in the context of all that structure. You would need to read carefully to catch how often they are glib or bullshit ideas. Bullshit can trick people, especially tired or disengaged people. Worse, as I said earlier I find my mind resists engaging deeply with AI-generated content. Carefully structured documents currently require a lot of work to create and therefore are a gate (for good or bad) that people need to pass through. In a world where in a few minutes anyone can produce a beautifully formatted 20 page project plan that references organizational values and policy, will our social structures (committees, etc) be able to disambiguate what is actually good from what just looks good?