Have we made a unicorn? Continuous SVG-pelican style benchmark (havewemadeaunicorn.com)

In a unique challenge within the AI and machine learning community, various advanced AI models have been tasked with the artistic endeavor of drawing a unicorn in Scalable Vector Graphics (SVG) format. For ten consecutive days, these attempts have been judged by a human expert, and the outcome remains consistent: none of the generated images have convincingly replicated a unicorn to the extent that they could be mistaken for human-created art. This ongoing challenge highlights both the creative limitations and the current capabilities of AI in artistic generation. The significance of this competition extends beyond mere entertainment; it calls attention to the evolving benchmarks for AI creativity and the push for models that can produce indistinguishable art from human artists. Models like Anthropic's Claude and various iterations of OpenAI's tools are among those being evaluated, underscoring the competitive landscape in AI artistry. This endeavor not only tests the technical skills of these models in graphical rendering but also ignites discussions surrounding perception, artistry, and the future potential of AI in creative fields.