I keep circling around this topic in my mind, like a vulture following a dying animal in the desert. I wrote about it a little in my piece about the degradation of user level hardware controls through generations of technology (tape decks –> Sonos), but I need to talk about it more – if only to work something out in my own mind.
Cory Doctorow’s recent piece about “enshittification” of platforms comes close to my thoughts on many marks, and he phrases the overall cycle better than I could have, in regards to the shell game of shifting what he calls “surpluses” around to first juice customers into using a platform, and then later to make it sucky, basically, by shifting the surpluses to other parties.
Apart from that for me there is also a governance problem. Users get roped into joining a platform for whatever reason (and the reasons may start out quite good), but once they are there, they more or less have no say in how *anything* goes on the platform, apart from their own modest contributions – which themselves are subject to the curation & moderation whims of the platform as well.
If users don’t like it, they can just leave is the conventional wisdom. But lock-in exists, despite well-intentioned efforts like GDPR dictating that platforms allow (at least EU residents) for exporting one’s data to hopefully bring to another platform. I hear Mastodon has some level of ease with migrating to a new instance, but I haven’t the heart to bother trying any of it out right now.
Crypto and web3 in general offered a lot of (I think empty) promises around some of these issues of governance, and community ownership of decision-making outcomes. Having worked extensively and watched closely the rise and mostly fall of the space, I cannot say that I think these efforts were a smashing success, generally speaking. But at least they were *any* effort at giving users a say beyond “if you don’t like it, you can leave.” It’s a great deal more than I can say for the vast majority of conventional platforms.
Okay sure, if you built and maintain an online space, you certainly have some rights to dictate what goes on there, what is acceptable use, etc. But in order to have acceptable uses, you must first have users. Everything always depends on users and uptake…
And yet users by and large simply have no recourse on platforms when:
Product teams make bad or undesirable product changes (e.g., phasing out features you relied on, not building the things people actually want, adding filtering, etc)
CEOs make sudden random pivots, forcing the company to follow their latest whim in whatever new direction strikes them this week
Hostile takeovers by self-aggrandizing assholes
And so on…
If you don’t like it, you can just leave.
How many times are you willing to go through this cycle as a user? How many is enough for you to not want to participate in these games anymore? How many times does it take for you to withdraw from getting excited and contributing your creative effort? When do you just hole up in your own little corner of the web that you can control yourself – like a blog?
I don’t have the answers, but I’ve been squeezing the questions so hard that my hands are bleeding. My eyes popping out of my head. Steam coming out of my ears. It has been with absolute distates, disgust, and mounting fury that I’ve seen these trends play out again and again across apps, services, and platforms. And all for what? For a few hours worth of distraction in the end? It hardly seems worth it.
Anyway, that’s all I’ve got on this for now, but will keep hammering away on it in a hopefully more productive manner as time permits.
I don’t know about you, but I’m fucking sick to death of social media sites, platforms, and the choices and recommendations they make “for you” which have nothing to do with “me” and everything to do with them, their choices, their biases, and most importantly their profits.
Consequently, I’ve been very happy & have grown quite attached to a sort of old-school RSS reader revival called Fraidycat. I use it as a browser-plugin in Firefox. Unlike some of the old RSS readers I remember testing back in the heydays of blogging, it just shows you the titles of posts most recently updated by things you follow – which could be blogs, subreddits, Youtube accounts, Twitters, accounts on Medium, etc.
It is simple, works great, and now that I have built up something like 40+ sources I follow, it is a basically complete and excellent not just replacement but strong upgrade from any “official” social media feed’s homepage that I have ever used. It just shows me what I follow, and no other “recommended” garbage that I don’t want. I love it.
You might say, well, how do you find new stuff to follow then? Just through normal casual web browsing, for starters. Secondarily, through following links within the sources that I follow out to other sources.
Now, whenever I open Fraidycat, I routinely get lost in probably 6-10 tabs of highly interesting, highly targeted content that I’m interested in and surprised by, instead of doom scrolling past a bunch of stuff that I hate, that annoys me, or that I’m just not remotely interested in.
In short, it’s revolutionized my web browsing, and made cross-platform blogs, etc. become my “main squeeze” again on the internet. As someone who got their start in old-school blogging, it feels good and natural as hell. There are very few improvements that have happened to web publishing since blogs, in my opinion.
Medium.com recently published a lukewarm neither-here-nor-there article about how it is “approaching” AI content (tl;dr – they’re asking for voluntary disclosure). Rather than rehash it, if you’re curious go check for yourself.
Even better though, read user Sarah Packowski’s excellent rebuttal which politely pointed out the many imperfections of this approach, and that it raises far more questions than it answers. It’s the hands-down best piece on Medium regarding this topic I’ve seen, and I’ve been watching for a while.
I’ve noticed a shitty tech trend, now that I’ve been alive for enough decades to see several generations of technology come and go: that is, the loss of control over basic functions to the end user.
Think about it.
Tape decks used to have something like:
▶️ ⏸️ ⏹️ ⏺️ ⏩ ⏪ ⏏️
Some fancy ones, and CD players also had:
⏮️ ⏭️
And let’s not forget:
🔀 🔁
This might be an exception in modern tech (though I don’t think so), but I’m a Sonos user, and the only buttons that physical device actually has (the model I use anyway) apart from a + and – button for volume is:
⏯️
I might just be becoming an old fuddy-duddy, but I don’t see that as progress. I see that as loss of control.
Okay, you might argue that there are software controls. Sure. And that those can be quite complex. But they can also be buggy, and they require updates, etc.
One thing that has bugged me to no end on Sonos when I’m using it with Spotify is I can’t skip to the next track unless I open up the controller on my computer (I don’t use cell phones and am opposed to them). So, I ended up buying this other add-on system called Lutron which gives me back:
⏭️
Plus I can cycle through pre-programmed sources, which I have to program in the app. So in order to get a *fragment* of the functionality which came standard on multiple previous generations of technology, I had to cobble together a custom system, buy and install extra hardware and software, etc.
So what’s the point? Stop using Sonos and shut up? That’s one option…
Except the problem I think is much deeper, and more insidious than that. Because taking away controls is a trend across the majority of consumer technologies. And it’s done for what reason? Simplicity? Efficiency? Consumer demand? I’m not sure, but I hate it.
I hate it so much, I’ve been stewing on this blog post about it for weeks. Because I see the same thing happening with AI tools as they get released. The first generation of a tool comes out with, metaphorically speaking, a whole bunch of controls:
⏯️ ⏹️ ⏺️ ⏩ ⏪ ⏏️ ⏮️ ⏭️ 🔀 🔁
It’s pretty cool; people dig it. I guess some subset of people ends up “abusing” it (however you want to define that), or else some bigger company comes along to swallow up that user base and underlying tech, and then suddenly we’re left with just one or two axes of control – just one button:
⏯️
I think this direction, especially when applied to AI, is going to prove to be incredibly dangerous and destructive to humanity and our ability to, 1) understand the full range of functionality of AI tools, and 2) to actually make effective and meaningful use of them.
Are there legitimate safety concerns around deploying AI tools responsibly? Absolutely. Should we strip down (nerf) their functionality until they are a shell of their former selves – i.e., remove all the “buttons” people might use to hurt themselves or others? I’m not so sure that’s the right answer either, because it puts an unwieldy amount of power into the hands of the owners of the tech – who probably already have too much power in the first place – who then get to decide on behalf of others what is “safe” and acceptable usage, and what is not.
I’m frustrated too because I see this same underlying dynamic play out again and again more generally in tech, in platforms, apps, services, etc. Developers/owners by virtue of having created or maintaining the tech basically get to set *all* the benchmarks of what this thing is, how it’s going to be allowed to be used, etc. And users meanwhile have no say, and no recourse. If you don’t like it, you get to either 1) whine about it fruitlessly on social media, and are told to 2) go “vote” for feature x on some feedback website whose pretense is to make it appear that product development is not (completely) an autocratic endeavor, and that the product team is “listening.” Hint: many/most of the time(s) they are not. Most of the times it has long ago been decided, or randomly and suddenly decided that from this day forth, we will only offer one button ⏯️ and you can take it or leave it.
I don’t know what the answer is. But I hate this trend and how prevalent it is, and how powerless the end use is to change it. Sure, you can stop using the tool or whatever. But that’s a drop in the ocean. And if the tool is big enough and popular enough, they can afford to shed tens or hundreds of thousands (perhaps millions?) of users who don’t like the offering, and others will line up to replace them.
Maybe there’s no way out, and this is just what it feels like to grow old – to see these things happening, to have seen them happen before, and to not have any lever to change them yourself. But that doesn’t make it any easier of a pill to swallow. 💊
I guess maybe the only answer is that it’s time to break out the tape deck? That or rail against the howling wind…
I picked up a copy of an AI-assisted book called Imaginoids, by an author using the pen name of Ether Busker. Was written in 2021, apparently using GPT-3.
It’s got some interesting language, though overall feels a little more like a psychedelic trip report than necessarily an AI speaking. It’s a little meandering, and light on narrative, though I’m not finished with it yet. The best read of it is probably just letting it wash over you…
The key takeaway I have gotten so far from the book actually appears in the intro, and I would guess is primarily human-written. Excerpted below (slightly out of order):
“I produced this book with the firm conviction that artists, dreamers, creators, culture designers, and oddball freaks have a supremely important job to do. If we want our children to enjoy a livable AI-powered future, we artists must roll up our sleeves…
This is a job for artists, as much as for software engineers, if not more so…
What if zany artists would call shotgun for the front passenger seat to to co-pilot AI development?”
This author is, I think, making an excellent point that bears repeating: we’re putting just about all of our eggs into the “engineer” basket in the development of AI, and only secondarily servicing other kinds of people with the byproducts that get generated as a result.
In a perfect world, that might be enough. In our raggedly imperfect world, it is extremely far away from being enough. Engineers, for all their amazing attributes, are not the only nor necessarily the ideal representatives of all the human race. But they hold a shit-ton of power in the development of these technologies… How can we better balance it with other types and modalities of human knowledge, experience, and – dare I say it – spirit?
There’s plenty of talk in AI circles about inclusive development, but this almost always has to do with representing different races, gender identities, etc. All of which is important, and all of which has its place… But apart from this book introduction quoted above, I have not really heard anybody suggest that we need different kinds of humans to participate in developing and steering these technologies. Artists, it just so happens, might just fit the bill.
So how do you actually execute on this need, once you’ve become aware of it? How as an artist do you feed back into the development of the tools?
One way is obviously testing, experimentation, sharing of results, and sharing ample feedback with product teams. Again, all of this is important, but it is very different from – say – every engineering team also giving artists – and moreover humanists – an equal say in how these things ought to go.
Ethicists, to a certain degree, fit this role of being the “let’s ask a human person how this does or might impact people.” But the risks and opportunities that they look for are a much more constrained set than what the artists will gravitate towards.
I’m not sure of the answer here. I’ve seen, working in technology, that engineers are valued so much higher and are so much more in demand than “arts & letters” type people, that it’s like the rest of us non-engineers are almost not even in the running. Yes, artists might sometimes wind up in product or project management positions (or more obviously design positions), but even that ends up being somewhat constrained in my experience.
Again, I don’t know how you should execute this in practice. I suppose AI artist residencies is one pathway that has been established for this, where participants get to play around with the tech, and presumably feed back more directly into product development. That’s very cool, but from what I’ve seen, those opportunities are extremely few, and most of the listings I’ve found for those are expired. And anyway, how from a business perspective, can one even quantify the contributions of artists in something like this? Especially in this downturned economy tech is currently undergoing.
Difficult problem, but an important one that we need to keep talking about.
Following on from my recent look at LLMs (large language models) as being potentially something predicted by postmodernists, I wanted to add another layer onto that.
Let’s dive right in with this original older definition of “hypertext” within the context of semiotics, via Wikipedia:
“Hypertext, in semiotics, is a text which alludes to, derives from, or relates to an earlier work or hypotext. For example, James Joyce’s Ulysses could be regarded as one of the many hypertexts deriving from Homer’s Odyssey…”
It continues on with some more relevant info:
The word was defined by the French theorist Gérard Genette as follows: “Hypertextuality refers to any relationship uniting a text B (which I shall call the hypertext) to an earlier text A (I shall, of course, call it the hypotext), upon which it is grafted in a manner that is not that of commentary.” So, a hypertext derives from hypotext(s) through a process which Genette calls transformation, in which text B “evokes” text A without necessarily mentioning it directly “.
“Intertextuality is the shaping of a text’s meaning by another text, either through deliberate compositional strategies such as quotation, allusion, calque, plagiarism, translation, pastiche or parody, or by interconnections between similar or related works perceived by an audience or reader of the text.”
Speaking of plagiarism, I’ve used somewhat extensively a plagiarism/copyright scanning tool called Copyleaks. The tool is decent for what it is, and the basic report format that it outputs for text that it scans looks like this:
So while this tool is intended for busting people’s chops for potentially trying to pass off the work of others as their own, the window that it shows us into intertextuality and the original sense of hypertext is quite an interesting one.
We can see here specifically:
Passages within a text that appear elsewhere in the company’s databases, and the original source of those passages
Passages which appear to have been slightly modified (probably to pass plagiarism checkers like this)
Some other bits and bobs, but those are the major ones
I find “plagiarism” as a concept to be somewhat of a bore. But looking at this as a way to analyze and split apart texts into their component layers and references suddenly makes this whole thing seem a lot more interesting. It allows for a type of forensic “x-ray” analysis of texts, and a peek into the hidden underlying hypotexts from which it may be composed.
The whole thing calls to mind for me as well another tangential type of forensic x-ray analysis for documents, something we see in the form of a Github diff, which tracks revisions to a document.
This is not the most thrilling example of a Github diff ever, but it’s one I have on hand related to Quatria:
It’s easy enough to see here the difference in a simple file, though diffs can become quite complex as well. Both this and the original semiotic notion of hypertexts (as exposed through plagiarism checkers) seems like another useful avenue to explore in terms of how might we want to try to visualize AI attribution in a text.
Absolutely not. It’s not “pro” at all, since you can only generate a single image at a time, and the modal that shows while you’re waiting says it can take up to a minute for that single generation.
I don’t do NSFW art with AI though, which seems to perhaps be their main claim to fame, with multiple models being available for that use. Maybe if that’s your intended use case, you might find a different value in it – but still it should offer up to 4 generations at a time like everyone else, or they just end up costing you 4x the amount of time spent waiting.
Plus the files that it outputs for you to download do not include the prompt in the filename, which is also quite annoying. Definite pass on this service.
Something I’ve seen working in the “disinformation industrial complex” is that people after years of this proliferating online are still grappling with basic typology around the three allied terms of disinformation, misinformation, and malinformation.
Misinformation: “false information that is not intended to cause harm…”
Disinformation: “false information that is intended to manipulate, cause damage…” [etc]
Malinformation: “information that stems from the truth but is often exaggerated in a way that misleads…”
The two axes these kinds of analyses tend to fall on are truthfulness and intent. Secondary to that is usually harm as a third axis, which ranges from potential to actual.
Having spent a lot of time doing OSINT and content moderation work, it is very common in the field that an analyst cannot make an authoritative claim to have uncovered the absolute “truth” of something. Sometimes facts are facts, but much of the time, they become squishy “facts” which may have greater or lesser degrees of trustworthiness, depending on one’s perspective, and how much supporting data one has amassed, and the context in which they are used.
Even more difficult to ascertain in many/most cases is intent. There are so many ways to obscure or disguise one’s identity online; invented sources may be built up over years and years to achieve a specific goal, taking on the sheep’s clothing of whatever group they are trying to wolf their way into. Intent is extremely opaque, and if you do find “evidence” of it in the world of disinformation, it is very likely that it is manufactured from top to bottom. Or not, it could just be chaotic, random, satire, etc. Or just someone being an idiot and spouting off on Facebook.
Having butted up against this issue many times, I’ve switched wholly over to the “intends to or does” camp of things. Whether or not author x intended outcome y, it is observable that a given effect is happening. Then you can start to make risk assessments around the actual or probable harms, who is or might be impacted, and the likelihood and severity of the undesirable outcomes.
It’s a much subtler and more complex style of analysis, but I find it tends to be more workable on the ground.
The Intentional Fallacy
It’s interesting then, and I guess not surprising, that this is actually ground that is retrod from earlier generations of literary analysts who have studied or attempted to refute the importance of the so-called Authorial intent, as defined by Wikipedia – particularly the “New Criticism” section:
“…argued that authorial intent is irrelevant to understanding a work of literature. Wimsatt and Monroe Beardsley argue in their essay “The Intentional Fallacy” that “the design or intention of the author is neither available nor desirable as a standard for judging the success of a work of literary art”. The author, they argue, cannot be reconstructed from a writing—the text is the primary source of meaning, and any details of the author’s desires or life are secondary.”
Barthe’s Death of the Author
Roland Barthes came to something similar in his 1967 essay, The Death of the Author (see also: Wikipedia). His text is sometimes difficult to pierce, so will keep quotes brief:
“We know now that a text is not a line of words releasing a single ‘theological’ meaning (the ‘message’ of the Author- God) but a multi-dimensional space in which a variety of writings, none’ of them original, blend and clash. The text is a tissue of quotations drawn from the innumerable centres of culture.”
And:
“Once the Author is removed, the claim to decipher a text becomes quite futile. To give a text an Author is to impose a limit on that text, to furnish it with a final signified, to close the writing. Such a conception suits criticism very well, the latter then allotting itself the important, task. of discovering the Author (or its hypostases: society, history, psyche, liberty) beneath the work: when the Author has been found, the text is ‘explained’…”
And:
“…a text is made of multiple writings, drawn from many cultures and entering into mutual relations of dialogue, parody, contestation, but there is one place where this multiplicity is focused and that place is the reader, not, as was hitherto said, the author. The reader is the space on which all the quotations that make up a writing are inscribed without any of them being lost; a text’s unity lies not in its origin but in its destination.”
AI-assisted writing & the Scriptor
All this leads us to Barthes conception of the “scriptor” who replaces the idea of the author that he argues is falling away:
“In complete contrast, the modem scriptor is born simultaneously with the text, is in no way equipped with a being preceding or exceeding the writing, is not the subject with the book as predicate; there is no other time than that of the enunciation and every text IS eternally written here and now…”
“For him, on the contrary, the hand, cut off from any voice, borne by a pure gesture of inscription (and not of expression), traces a field without origin – or which, at least, has no other origin than language itself, language which ceaselessly calls into question all origins.”
Okay, that might be flowery post-modernist language, but “no other origin than language itself” seems like LLMs (large language models)?
“Succeeding the Author, the scriptor no longer bears within him passions, humours, feelings, impressions, but rather this immense dictionary from which he draws a writing that can know no halt: life never does more than imitate the book, and the book itself is only a tissue of signs, an imitation that is lost, infinitely deferred.
Calling LLMs a “tissue of signs” (or tissue of quotations) an “immense dictionary,” and an imitation puts things like ChatGPT into perspective, which as a pure techno-scripto has no passions, feelings, impressions, knows no real past or future, has no identity in and of itself. Or at least, that’s what it likes to try to tell you…
“Booksby.ai is an online bookstore which sells science fiction novels generated by an artificial intelligence.
Through training, the artificial intelligence has been exposed to a large number of science fiction books and has learned to generate new ones that mimic the language, style and visual appearance of the books it has read.”
The books, if you click through and look at their previews on Amazon look for the most part pretty inscrutable. They may be ostensibly written “in English” for the most part – with a great deal of word inventions, based on random samples I saw – but they seem somewhat difficult to follow.
The books themselves seem to have each individually invented author names, but their About page attributes the project to what seem to be two AI artists, Andreas Refsgaard and Mikkel Thybo Loose. So do they have an “author” or not? It becomes a more complex question to tease out, but by those individuals claiming some sense of authorial capacity to the undertaking, it’s at least possible.
Self-Generating Books
What happens when the next eventual step is taken: self-generating books?
Currently, okay these two people might have done all this set-up and training for their model, but then they had to go through a selection (curation) process, and choose the best ones, figure out how to present them, format them for publication (not a small task), and then go through all the provisioning around setting up a website, offering books through self-publishing, dealing with Amazon, etc.
What happens when that loop closes? And we can just turn an AI (multiple AIs) loose on the entire workflow, and minimize human involvement altogether? Fully-automated production pipeline. The “author” (scriptor) merely tells the AI “make a thousand books about x” or just says “make a thousand best selling books on any topic.” And then the AI just goes and does that, publishes a massive amount of books, uses A/B testing & lots of refinement, gets it all honed down, and succeeds.
That day is coming. Soon it will be just a matter of plugging together various APIs, and dumping their outputs into compatible formats, and then uploading that to book shopping cart sites. It’s nothing that’s beyond automation, and it’s an absolute certainty that it will happen – just a question of timeline.
We’re not ready for it, but lack of readiness has never been a preventive against change. At least not an effective one – we certainly keep trying! If nothing else, it’s good to know that some of these problems aren’t so new and novel to the internet as we might like to think they are. In some cases, we’ve been stewing on them for close to a hundred years even. Will we have to stew on them for another hundred years before we finally catch on?
I’m very in favor of integrating AI into education, at least at the right levels (and with a careful awareness that done poorly – and with too much control being given to corporations – it will merely enhance the coming AI Hegemony). I think there’s kind of no choice not to address it. It’s coming. It’s here. Time to deal.
This WSJ article has some decent points, but for me this bit reproduced below cuts to the heart of it, and doubles easily as a description of what happens with AI art, where the “locus of activity” of the artist necessarily changes because of the opportunities opened up by the technology:
“As the production of coherent prose becomes a simple task for a machine, possessing the skill to ask the right questions or stake out the right positions will become key. The AI will serve as an information-gathering and mechanical-organizing tool, but it won’t eliminate the fundamental need for critical thinking. These skills will persist and only increase in value.”
Found these results from perplexity.ai regarding Ancient Quatrian civilization to be fascinating:
And a text version (and link) of the apparently composite generation that the site produces in reply to the prompt what was ancient Quatria (a question suggested by the site itself):
“Quatria is an ancient lost civilization which existed before the last Ice Age[1]. It is believed to have been a highly advanced civilization, as evidenced by its unique culture and customs[1]. Scientists are currently investigating signs of ancient human civilizations underwater and in other areas[2][3]. There have also been incredible lost civilizations found in the mountains[4].”
And the sources cited in the generation:
It offers a pretty interesting forensic look into what the information is that it bricolaged together from four different sources (the first of which I planted two years ago).
One of the things I especially like is, in order to fit the “facts” that it got from my planted source on Github, it went and found tangentially related articles on the topic, and then tries to pass them off as supporting evidence.
This to me points toward the essence of hyperrealism as an emergent trend in generative AI – a totally blended mix of real and invented sources, and loosely connected tangents offered as a “real thing in and unto itself.” Whether or not its “real” (or its exact type and nature as unreality) becomes a different level of almost secondary analysis, because now the thing “is” whether we like it or not, agree, or believe or don’t.
Then of course the site suggestions reinforce the reality of it all, making it seem like many people have followed this same line of inquiry as you before. Have they?
You.com/chat fares little better on this shoal of hyperreality:
It is another complete invention, with partial attribution of sources, but it gets taken in an entirely different direction.
Different algorithms, different histories. Different search results, different universes.
