Does anyone else feel like generative AI has plateaued?  Remember 2022, when text-to-image generators like  Midjourney and DALL-E debuted? I think we all contributed to melting a glacier or two with our gleeful exploration of these  platforms. Once I mastered writing prompts that resulted in  images beyond the realm of "uncanny valley meets ’80s airbrush  poster art"—an aesthetic AI favors—I was enamored. My prompts became consistent enough that Denver-based artist  Jared David Paul asked me to collaborate to create prompts  describing sculptures he’d made. The final AI-generated visuals were printed and exhibited alongside the sculptures. They were  strange and imperfect, but there was an undeniably compelling relationship between the physical objects and the AI-produced  images. 

Then, in late 2022, ChatGPT, a large language model (LLM),  was publicly released. Suddenly, the future—for better or worse —had arrived. Schools banned AI, and headlines warned of  doom for copywriters and novelists. 

With 22 years as a visual designer, art director, and animator,  staying relevant is an important focus for my career. Rather than  waiting for AI to make me obsolete, I dove in. I’ve found AI  useful for planning a trip to Alaska, retouching images in  Photoshop, rapid wireframing in Figma, and refining work  documents. In 2023, I enrolled in "Artificial Creativity" at  Parsons School of Design. There, I learned about the historical  foundations of human creativity and how AI might succeed or  fall short in replicating these traits. I have embraced AI as a tool,  accepting that it could eventually mean the dissolution of my  career and many others. I hold out hope that, as taxpayer bases diminish, governments will enact laws to protect workers or offer universal basic income (UBI) so we can be unemployed without becoming unhoused. Little attention is paid to workers losing their jobs, but attention will be unavoidable when no taxes are being collected. 

I am writing specifically about the generative AI and LLMs proliferating in the creative industries. I know very little about AI for medical or manufacturing purposes for instance and would not want to make statements that I cannot support. I believe those tools as well as the “agents” many tech companies are currently working on are indeed advancing at an impressive  rate. The tools I am most familiar with are Midjourney, DALL E, ChatGPT, Claude and Adobe’s Firefly, which can be found  natively in Adobe creative apps. I’ve been exploring DeepSeek.  Its UI is intuitive, and it seemingly matches ChatGPT’s capabilities while being open-source, 30x cheaper to train, and  95% less resource-intensive. I’ve even gone so far as to build  my own GPT. I trained an AI career coach on Machiavelli’s The Prince. War and business aren’t so different!  

Once the knowledge base is there, it doesn’t devolve. Unlike  humans, AI won’t forget something it learned before. These  technologies haven’t slipped since they were first released, but  my question to you is, when using AI, are the results reliable? 

Recently, I asked Photoshop to remove an image background.  Instead, it created a new one featuring a woman beside a robot. Illustrator fared no better when I requested vector art of a motmot bird—a well-documented species. Twenty seconds later,  I received a black-and-white image of a man in a suit. This isn’t  useful or quirky—it disrupts my workflow. I ended up drawing the bird myself. I probably prefer it that way. I don’t want to be replaced or see the industry flooded with "designers" who are just prompt engineers. 

For a presumably pro-union audience, that sentiment should  resonate. I asked ChatGPT why visual AI lacks the sophistication of language models. It responded with two self congratulatory paragraphs—AI has learned ego from us too! But its key point was: "Unlike text models, which interpret language,  visual AI must grasp complex elements like color harmony,  spatial balance, and stylistic consistency." Replace "visual AI" with "designer," "illustrator," or "photographer" to see why we  remain essential. Writers, I haven’t forgotten you. Great writing  captures human depth—emotions, challenges, dreams. AI  writing B2B blog copy? Fine. AI writing novels? I’ll suffer a small death. 

Frankly, I’m relieved AI adoption has been slower than predicted. The tech industry, true to form, pushed unfinished products to market. "Move fast and break things"—what a  model! No doubt, tech CEOs and venture capitalists (hoarding  supervillains) made obscene amounts of money in the process, but they may also have unintentionally provided us with a delay that will help protect jobs and incomes. AI still lacks nuanced  human abilities, keeping us competitive. I’m no Luddite—AI is here to stay. Master relevant tools. Work faster but charge the  same. This window won’t last forever. Call it "greedflation" compensation.

If you aren’t already, get involved. Shape the conversation.  Contact representatives, join the Freelancers Union, and discuss AI with your employer. We must push for protections in creative work, ensuring fair compensation for human labor. Build unique  skills that differentiate you from machine-generated work. AI  and humans can coexist, but we must foster a market where authenticity and originality are prioritized. Together, we can lead the way for the responsible integration of AI and create an  industry where technology complements, rather than replaces,  human creativity.