 |
| Picture Limitless Creativity at Your Fingertips |
A few days later I'm reading WIRED and Kevin Kelly nails down not only a way to migrate into this disconcerting future full of AIs, but also how it works and why everyone shouldn't be terrified of it:
"Generative AI will alter how we design just about everything. Oh, and not a single human artist will lose their job because of this new technology."
That's a pretty daring prediction, but Kelly goes on to explain that most AI generated visual art isn't being used in traditional places. Instead it's a deeply personal form of visual expression hyper focused on the co-creator (the human working with the AI). As a result, much of the art produced by AI acts as a kind of art therapy. That hyper personal focus is something education systems (based on 19th Century factories) really struggle with, but it's the future if honouring equity and diversity matter in any real way.
What has me most excited about AI is precisely that it treats people more like individuals than people do. If a future education system has AI teaching assistants that radically individualize instruction, then I consider that a massive win for children and a huge step forward in terms of pedagogical best practice. The only resistance to this individualized, equitable technology enhanced future would be by the people managing the current status quo for their own benefit.
Instead of a teacher who barely remembers your child's name, let alone his reading level or his neurodiversities, an AI that responds to his needs precisely and when he needs it to sounds like the kind of future I want to live in. We're currently happy with loud attention seekers getting most of the attention in class while the quiet ones fall through the cracks, and by 'we' I mean the people running public education.
In the article, Kelly goes on to talk about how AI democratizes visual arts. There is a follow up piece that talks about the 19th Century panic around the emergence of photography (painting is dead!). Each of these technology disruptions put visual mediums into more people's hands. There was a similar panic when smartphones made everyone a photographer in the early 21st Century. Just because you have a camera in your hands doesn't make you a photographer, any more than waving a paint brush makes you a painter. There is intentionality implicit in art that technical skill is only a part of. Machines that make it more accessible don't give the users those esoteric art skills needed to make art, but they do offer a 'lowercase' opportunity to visually express your ideas.
The AI in Your Classroom presentation Heidi and I put together gives examples of how two teachers focused on evolving pedagogical best practices have explored AI. In my case we got into it five years ago when a parent who happened to be on the IBM Watson team offered to get us going on coding AI powered chatbots. We found this so easy to do that it ended up being part of our grade 10 coding curriculum along with web development. At the same time my game development students were discovering that we needed to get a handle on enemy AI scripting in Unity or our games wouldn't play well (stupid enemies are boring). The introduction to AI had students learning about intents and thinking about coding in a collaborative rather than dictatorial way. This not only enriched our game development scripting, but it also led to seniors exploring adaptive algorithms in large data-sets and exploring GitHub's CoPilot. This approach has had us learning 'AI Whispering' years before it became an emergent media literacy...
"Behind this new magecraft is the art of prompting. Each artist or designer develops a way of persuading an AI to yield its best by evolving their prompts. Let’s call these new artists AI whisperers, or prompt artists, or promptors. The promptors work almost as directors, guiding the work of their alien collaborators toward a unified vision. The convoluted process required to tease a first-rate picture out of an AI is quickly emerging as a fine-art skill." - WIRED
Reading Kelly's article focused on AI visual design a week after doing that AI in the classroom presentation resonates. The one time I was allowed to teach visual arts in high school, I was given a grade 9 curriculum that was so prescriptive that the projects in it sounded like directions for an AI: 'Make an acrylic painting using the techniques practised in class in the style of the Group of Seven'. 'Using perspective drawing techniques, draw a picture of a room in your house'. I understand the need for these skills focused assessments, but the opportunity to create wasn't something grade 9s were being encouraged to do. The entire course was entirely focused on lowercase creativity making it very easy for an AI to do the work. Maybe the seniors got to explore uppercase creativity, but I never got a look in at any of those classes. This raises difficult questions around how we're going to develop skills when AI keeps stepping in to do work that students only see as busy work.
The other thing this visual arts department fixated on was photography, which they described as 'real' photography complete with dark room and chemicals. The follow up WIRED art history piece on the 19th Century photography panic casts a poor light on that thinking. Photography was considered the 'end of painting' when it first appeared, but nothing of the sort happened. What the new medium of photography did was create new influences that enriched both painting and the emerging medium. AI visual design will likely follow a similar path. What doesn't work is clinging to an old way of thinking rather than encouraging this enriching influence.
 |
| Dall-E AI generated future makerspace |
I found my way back into photography when we lived in Japan after I graduated from university with massive student debts (another economically dictated decision). Digital cameras were just coming out in Japan in the late 90s and I got my hands on them early. This technology democratized photography and let me get back into it without bankrupting me. Hearing an arrogant gatekeeper-to-the-arts telling me that what I did wasn't real photography was difficult to take. This approach to the arts is quite common in high schools.
 |
| Dall-E AI Generated: future classroom |
So much technology use is hobbled by the politics of education. Black and white politicizing means unions paint technology as teacher replacing evil, which gives politicians interested in pushing education into private hands exactly what they need to do that. When no one is focused on improving pedagogy, we all lose. What education should be focused on is using every means at out disposal (technological, social or otherwise) to individualize and amplify student learning. Sometimes that might mean technology supplants a teaching position, but more often it should mean we are reorganizing how we do things to produce better learning outcomes though intelligent use of emerging technologies.
AI will certainly shake the foundations of education. The dinosaurs fixated on taking us back to the good 'ol days of rote learning and mono-cultural absolutism where all students learn rigorously in the same way should be far behind us, but those are the days every anti-technology educator longs for - when things (like photography) were 'real' and inaccessible to the proles.
The thing I'm most looking forward to with AI in education? It so quickly makes what we're doing in our low resolution factory-designed classrooms look poor that it will turn over the apple cart and force public education to reboot into a more individually focused, student-success driven model. In those future schools students will experience mastery learning based on their own abilities and will be supported individually. Some students may graduate high school at 15 years old, others may need it until they are 20. Learning would be hyper personalized and teachers would be conducting an orchestra of supports and data rather than overseeing a low-resolution classroom where individuals matter less than their age.
 |
| Education will focus on making this not true for as long as it possible can, but it shouldn't. |
From an artist's point of view, Kelly doesn't explain the place of human made art in a future where 'lower case', 'rote' art, ie: what is evidently taught in most schools, will be machine created. This raises some interesting questions about how we teach complex skill sets (like photography or painting) in a world where the skills are expressed by machines rather than people. If machines are reading to you and writing for you, do we need to know how to do those things either? Chasing that goal would lead to dire consequences. There are real benefits to human beings learning complex, tangible skills, both psychologically and economically.
When I take my decade old SLR out and catch a moment in nature because its talking to me, I'm not chewing through processing resources so that a machine intelligence with no understanding of nature creates an inherently abstract representation of what I'm experiencing. I take photographs to catch the light and make a moment permanent, so you can float in it. Technology assists me in doing that by helping me collect and share the data from that moment, but that data has a realness to it that no AI abstraction, no matter how well generated, will manage - because the AI and the media it creates is implicitly not in the world experiencing it directly.
 |
| When I was out in the polar vortex, I discovered that the ice crystal structures forming were incredibly complex compared to what I usually see, so I started working the macro lens as though I were seeing Joan Miro paintings. Imperfections and unconscious details render a natural truth in the photograph. Technology assists, but doesn't replace the subject. |
 |
| Dall-E understands macro photography and ice and creates a credible copy, but its always going to be inherently ethereal because it is an abstraction rather than a moment. Even well worded, maximum-AI outputs are going to lack an inherent realism until the AI itself is put into contact with nature directly - which too will eventually happen! |
I'm also not tied to the internet while I'm out in the world photographing. At its very best, an AI might be able to imitate getting close to nature, but it could never actually do it because of what it is. In the same way that a spectacular CGI effect in a film is almost too realistic, a good old-fashioned stunt with IRL effects offers nuance that makes it feel more genuine (because it is). No matter how good AI gets, it'll always be imitating that immediacy until it gets to experience it directly and then express its understanding of that experience.
 |
| Is having machines do it all for you really good for you? |
