Other – Page 79 – Tim Boucher

Art school

On 23 December 2022

As an art school drop-out, this quote about AI art models from illustrator James Gurney (Dinotopia) hits home:

“How do you run an art school when the students can solve any assignment with the push of a button?”

Thanks to ChatGPT this problem doesn’t apply merely to art schools, but more broadly to all education.

Art school is a particularly good test-bed though, because “art” can mean so very much. Honestly, I’d love to see art schools directly address this exact question. Art is so so much bigger than “making a picture.”

To me, the “problem” is not that much different than what artists encounter when they need to collaborate with other artists, which is especially apparent in theatre (where I worked for some 6 years). Only there the problem is somewhat clearer, since the issue is not merely about pressing a button and having a result, but working with a partner who is in control of the result (which, to be fair, Gurney addresses working with AI tools as partners). How do you exist as an artist when you have to rely on another person to produce a result?

Obviously, relying on another human being vs. relying on a computer for a result are two qualitatively different things. But perhaps that distinction won’t last all that long at the current rate of progress.

One thing I saw in art school that was never really overtly addressed anywhere was this underlying notion of the artist as the pinnacle of the rugged individualist. Artists seeking their own style, their own self-expression, their own fame, etc. Perhaps you might engage in the occasional collaborative or group project, but the focus is very much you you you.

Perhaps this is linked somewhat to fears that some artists seem to be facing around AI tools, that they feel there is a diminishing of the you you you of the rugged individualist artist with their own “unique” style etc. etc. A dawning recognition that, hey, we’re all basically doing more or less the same shit in more or less the same mass cultural context. That we’re even – gasp – by and large interchangeable.

For me, that’s not scary. That’s a relief. It means that to some extent I can shelve myself & my baggage as primary concerns, and merely execute within a cultural context as one of many exemplars. This accords more with past artistic traditions (I’m thinking medieval art esp.), where an artist was recognized as a link in the chain of tradition passed down through the culture. It wasn’t necessarily desirable in those traditions for artists to have their own unique style, so much as to transmit the style carried by the tradition.

Anyway, I don’t have all the answers, just lots of thoughts…

ChatGPT conspiracy disclaimers

By Tim B.

On 21 December 2022

In Other

Am asking ChatGPT for some help researching Project Blue Beam, and keep seeing it say variations of this:

“It is important to remember that conspiracy theories like Project Blue Beam are not based on verifiable facts and should not be taken seriously.”

I think this approach to “safety” is well-meaning, but potentially misguided. Shouldn’t we take conspiracy theories seriously when they have real-world implications? That seems like a no-brainer. Not taking them seriously is what leads to things like armed insurrections.

I’m not sure what would be the “right” answer to give here, but this to me doesn’t instinctively feel like the one that will persuade people to suddenly give up their views.

For me, there are shades of the below video concept here, of chastising users in a way that is probably unhelpful:

Not hiring artists

By Tim B.

On 18 December 2022

In Other

One of the major tropes I’ve seen around generative AI art is that it is already or going to soon take jobs away from artists.

I suppose it may be true in certain circumstances, but for my use it is simply not so. That is, I would never be able to afford hiring dozens of artists to produce hundreds of images to use in ebooks. It’s not a question of whether or not I respect the work/effort/learning etc of artists, because I am one. I just don’t have that kind of money for fun small side projects, nor do I have the managerial capacity to organize all the work of others that would require. So it’s not like, okay I have $50K to spend on art, that I am now not spending.

If anything, the only artist I am potentially replacing is myself, because in all likelihood, I would be the one executing on a small fraction of the image ideas that I have. And if I was lucky and had a ton of free time to commit, I might be able to push out say six decent quality images for a book in a few weeks. And that would still be an accomplishment. But with these tools, I can blast out ideas and images that allow me to construct my interior worlds externally. I can achieve things an order of complexity greater than what I could by hand. And that is exciting as shit, whether or not other people are happy about it.

How to feel

By Tim B.

On 15 December 2022

In Other

I’ve seen this argument going around a lot about needing to label fiction as fiction on web platforms. So I guess as not to confuse people… There are a lot of things wrong with this, all of which I’d like to unpack eventually.

But one of which I finally put my finger on: telling people how to feel. It’s annoying.

And this is effectively what forcing someone to label something as “fiction” really does, is it sets up the reader in advance (if they even check the label) a framework for how they “should” feel about it, and how far they ought to take their “belief” before it becomes “dangerous.” etc.

The Onion also made an excellent case against being forced to label parody, which I won’t re-summarize here. More on this as my brain works through it.

Conspiracy as a speculative fiction genre

By Tim B.

On 14 December 2022

In Other

Been thinking about how conspiracy is a subgenre like any other niche, like vampire erotica or what-have-you. Each with its own stylistic forms, tropes, etc. and expectations on the part of readers about what it ought to contain, or at least play with.

There’s a quote on this Wikipedia page about the thriller subgenre, conspiracy fiction. It is not quite what I mean by just pure conspiracy theory content, but entirely relevant:

“The protagonists of conspiracy thrillers are often journalists or amateur investigators who find themselves (often inadvertently) pulling on a small thread which unravels a vast conspiracy that ultimately goes “all the way to the top.”

The difference between this thriller > conspiracy subgenre (as described in above quote) and that of conspiracy theory “proper” is that the protagonist of the latter is not to be found in the pages of the work, but is the reader, who becomes the amateur investigator.

This is also tied up in the format of modern conspiracy theory storytelling (across forums and platforms) as a type of networked narrative or transmedia storytelling. Where each reader has their own journey into and through the material via web artifacts encountered in particular contexts online.

Which is to say that, writing in this genre, it pays to emulate the format norms of making many diverse artifacts available, each one digging down into a specific topic, building in a web of cross-references.

In a way, structurally how conspiracy theories operate is extremely well-suited to hypertext in the first place, since short non-linear blurbs that are easily digestible tend to displace more meaningful and contextual longer works. Blah blah blah. It’s a blog post, it doesn’t need to make sense or have an ending.

Credit cards

By Tim B.

On 12 December 2022

In Other

Have a feeling that creators not wanting to use AI right now will look in the future the way that it looks now when people in the 90s were saying they would never use a credit card on the internet. Mmkay.

Find your demon

By Tim B.

On 5 December 2022

In Other

No matter what your demons are, you will find them in AI technology.

Information should be ‘dangerous’ (sometimes)

By Tim B.

On 3 December 2022

In Other

One thing I see a lot of lately is this idea that some information is fundamentally dangerous. I don’t have this all worked out yet, but wanted to challenge that base assumption, by unpacking a little what danger constitutes in relation to information.

“Dangerous” information would presumably be information that threatens some established conception of order. Let’s put aside for a moment which order is threatened by which information (and whether that order or that information is “legitimate), and try to understand the nature of a threat in this context…

With regards to information, it seems that a threat here might constitute a challenge or request to *change* a given conception of order. Depending on the circumstance, that may or may not be warranted or a wholly “good thing.” But danger is something that causes things to change: we need to run away to escape danger, or we need to actively neutralize the danger and contain the risk.

Is it possible then to have change in information systems without “dangerous” information that challenges a given order? Maybe. But that seems like it might be “boring” in many contexts (though “dangerous” information will always be unwelcome in some settings that require or at least heavily favor stability).

Will give this more thought and add on as I go, now that I’ve established a beach head here…

Interrogating AI models as an artist

By Tim B.

On 1 December 2022

In Other

Wrote this somewhere else that I can’t remember, but wanted to re-capture the thought here as I think it’s an important one.

In my mind, the correct way to use AI tools as an artist (or writer) is not necessarily (strictly) to use them to “create art.” Because anybody can do that, and get similar results, whether “artist” or non-artist… that’s the point, that they are democratizing technologies.

So instead, what feels right to me as an artist (ymmv) is to interact with the tools in such a way that you “interrogate” them broadly, deeply, and meaningfully. In other words, engage with them at a level that only an artist could or would think to do. Whatever that is will be different for everybody (and in the end, perhaps there is no such thing as a “non-artist” – but still). As them probing questions, challenge them on what they can do, explore the boundaries of the possible, and document it all alongside the reactions and impacts it all seems to have on real people.

That is, transcend the single-image output as your end product. Obviously, if you’re doing the above deep interrogation of the essence of these tools, you’ll generate hundreds or thousands of intermediate products (whether images or text) along the way. Let that become your trail, the record left in the wake of your passing through with honesty, curiosity, and creative rigor.