Blog/Book

April 2, 2024

AI Aesthetics and Art? A revolution? Or an oxymoron?

It seems some AI theorists are beginning to attribute an aesthetic to AI art generation, and finding lots to talk about in the process. I do experiment with AI in my work, and have quite a bit of respect for its mechanical capacities for production, generation and iteration. I think that AI does add something very powerful to artmaking as a form of amplifier and relational appropriator. Yet, I have determined that AI, when left to its own processes, rarely generates work that I am satisfied with. And the AI art that I've seen is not often satisfying or compelling either.  Can there really be a critical dialogue about 'AI's artistic ability' or 'aesthetic'? I decided to have a fundamental discussion about  this 'AI Aesthetics' with Google's Gemini. 

Dain Olsen: Is there an authentic "AI aesthetic" that can be discussed beyond the sophistication of its programming? Since aesthetics is neurologically based in embodied sensory experience and visceral meaning, can we truly attribute an aesthetic to AI? Just as with 'intelligence', I think we are talking about a machine's programmed capacity for mimicking sensory meaning, which doesn’t amount to 'meaning' as humans equate it with art.

Gemini: You raise a fascinating question about the nature of an "AI aesthetic." Here's how we can break it down:

The Challenge of Embodiment: You're right. Aesthetics are deeply tied to our embodied experience. The way light hits our eyes, the texture of paint on a canvas, the music that resonates with our emotions – these are all rooted in our biology. AI currently lacks that physical experience.

Mimicry vs. Intrinsic Understanding: Current AI art excels at mimicking existing styles and generating novel variations. It can be incredibly impressive, but it's like creating a collage of existing works. Can AI truly understand the deeper meaning and emotional impact inherent in human-created art? That's a question we're still grappling with.

The Emergence of Something New? However, there's another way to look at it. AI can explore vast creative spaces beyond human limitations. It can generate nonsensical yet strangely beautiful forms, or create music that bends the rules of harmony in intriguing ways.  This exploration might lead to aesthetics that are entirely new, ones we can't even conceive with our human-centric understanding of beauty. 

A Collaborative Future? Perhaps the most interesting aspect is the potential for human-AI collaboration. AI can be a powerful tool for artists, generating new ideas and pushing boundaries. The human artist can then infuse these creations with their own emotional depth and understanding, creating a truly unique aesthetic blend.

The "Meaning" of Art: Finally, your point about "meaning" is crucial. Art isn't just about mimicking sensory experiences. It's about provoking thought, sparking emotions, and offering new perspectives on the world.  Can AI achieve this? Maybe not in the same way humans do, but it could develop its own unique way of conveying meaning, one that challenges our current definitions of art.In conclusion, an inherent "AI aesthetic" in the human sense might be debatable. However, AI's role in creating new forms of beauty, exploring uncharted creative territories, and potentially co-creating with humans is an exciting prospect that could redefine our understanding of aesthetics altogether.

Dain Olsen: Everything you’ve said begins and ends with human intention and determination. AI has done nothing intentional. Correct? And certainly it is agreed that AI can push the parameters faster and further, although ultimately limited to pathways formed by humans. Real ‘collaboration’ then would be limited to similarly intentional, selective humans. With AI, it is human controlled from beginning to end. It is more like working with a lifeless dynamo which has its own gravitational force. 

Gemini: You make a strong point about the current state of AI and its role in aesthetics. Here's how to consider your perspective:

Human Control: The Puppeteer and the Dynamo: You're right, current AI art creation is very much under human control. We set the parameters, train the models, and ultimately interpret the results.  Your analogy of a dynamo with a gravitational pull is apt. AI throws out possibilities, and we, as the puppeteers, try to guide and shape them into something meaningful.

The Blurring Lines of Intention: However, the line between human intention and what AI contributes might blur in the future. As AI becomes more sophisticated, it might start introducing its own "mutations" into the creative process.  Imagine an AI that, after ingesting massive amounts of artistic data, starts generating works that challenge established styles or even propose entirely new forms of aesthetics.  Is this still just human intention playing out, or is there a hint of the AI itself shaping the outcome?

The Long Game of AI Evolution: Perhaps the "authentic AI aesthetic" you seek won't emerge overnight. It might be a long evolutionary process where AI, through trial and error, starts to develop its own internal representations and preferences.  We might be far from that point now, but as AI continues to develop, the boundaries between human control and AI's own contribution could become fascinatingly ambiguous.In essence, you're right that current AI art is human-driven. But the future might hold a more collaborative space where AI's own "voice" starts to emerge, subtly shaping the aesthetics we create together. It's an exciting yet thought-provoking scenario to consider.

Dain Olsen: You sound as though your programmers have somehow infused your responses with their own idealism regarding your prospects! When AI gains embodiment, somehow through some form of bio-technology, then it can possibly gain these attributes. But that only proves my point. This conversation seems to have confirmed the value of actual embodiment for forming aesthetic meaning. I wonder why we so easily invest in machines to make human products, rather than invest in our children's education and their aesthetic development. Imagine a world where we encouraged our children to make art! 

September 10, 2022

MY RESPONSE TO THE LA TIMES ON PANDEMIC LEARNING LOSS

The LA Times reported today that “L.A. student scores show deep pandemic setbacks, with 72% failing to meet math standards”. 

This was my Letter to the Editor (unpublished):

Dear Editor,              

Sadly, low test scores, particularly for the socio-economically disadvantaged, is not news. And it isn’t surprising that this was exacerbated by the pandemic and conditions of remote learning. According to the National Assessment of Educational Progress (NAEP), which has maintained a consistent standard for testing the breadth of academic proficiency for over 50 years, only about 37% of students are regularly proficient in literacy and numeracy, with significantly lower performance among the same students. The headline should actually be that this educational system is detrimentally flawed and intrinsically biased, apparently with little intent or understanding of how to fundamentally reform itself.

The fundamental problem is that this system is based in narrow, culturally biased assessment, which begets narrow forms of instruction, which are ineffective, passive, non-engaging, decontextualized and irrelevant, particularly for the socio-economically disadvantaged. A more advanced educational system, such as one centered around media arts education (https://www.mediaartsedu.org/), would measure individual student progress in rounded cognitive development, which would not be narrow or culturally biased. This would engender more effective and equitable forms of instruction, which are engaging, active, contextually rich, and relevant to all students and our current 21st century society. 

Dain Olsen

Co-Chair, Media Arts Committee, National Coalition for Core Arts Standards

Media Arts Teacher and Administrator, LAUSD (Retired)

Parent of LAUSD Students

July 25, 2022

WHY IS US STUDENT ENROLLMENT DECLINING?

Why are we seeing dramatic declines in student enrollment across the US? 1.3 million less students nationwide. New York and CA are both down 9%. You would think, as the pandemic wanes, we would be seeing the opposite, families clamoring to return to in-person learning. 

Where are all the students going? Apparently, one major reason is that many of these students have come to prefer online learning and homeschooling. This doesn’t seem to make sense, because the general sentiment was that remote learning was a disaster, both in its low appeal, and for student learning, which has set students back nearly a year in their proficiency. 

But from my perspective, and the perspective of Media Arts Education, it does make sense. During the pandemic, I believe that students and parents realized something important through remote learning. Remote learning wasn’t as far removed from the live classroom experience as we previously thought. For most core subject areas, the method is essentially the same - the teacher presents the lesson, models the learning activity, and then the student carries it out, with a bit of teacher monitoring. Thus, not a whole lot different, except that remote is through a video camera. The student doesn't have to be in the room in order to have this kind of lesson, as we see from online educational platforms, such as Kahn Academy and even Youtube.

In addition to its similarity to in person learning, online learning is safer, more convenient, flexible in scheduling, and seemingly just as effective, at least for parents who can support it, and students who are somewhat self-directed and motivated. 

Remote learning exposed the essence of our standard teaching practice, which is passive, not very interactive, engaging or even personalized. Students don’t need to go to school in order to experience that. Then, students may be asking, why go to school exactly? 

Schools should glean this lesson from the remote learning disaster by asking,  "What are we doing in classrooms that is greatly different from online learning, and achieves more meaningful and deeper learning objectives?" That is what we need to start offering in order to compete with online education and advancing education as a whole. If going to school was tremendously engaging, socially vibrant, meaningful and purposeful, then students would probably be much more enthusiastic to return to in person instruction. 

MAE can offer and support that kind of enriched learning experience and more robust learning outcomes.