top of page
Contact Me
HUMAN ALT-TEXT (2021)
"Human Alt-Text" takes a speculative approach, exploring the intersection of human labor and machine recognition through a performance-based framework. It draws parallels between how machines learn, analyze, and recognize images and how humans engage in similar cognitive processes through language and memory.
"Human Alt-Text" prompts reflection on the relationship between humans and technology. It challenges the audience to consider how we, like machines, process and interpret information. By engaging in this exercise, participants highlight the inherent labor involved in understanding, recognizing, and communicating knowledge — an act that is central to both human experience and machine functionality.
Through this speculative lens, the artwork invites viewers to rethink the roles of both humans and machines in the ongoing evolution of communication and understanding in an increasingly technological world. It poses a thought-provoking inquiry into the potential futures where human cognition and technology are intertwined.
> Performance Framework
* Methodology & Approach
Participants are tasked with memorizing a list of terms, each corresponding to a chosen image. The act of recitation serves as a form of script training, where individuals not only recall words but also translate them into descriptions, similar to how machines generate 'alt-text' for images. This process reflects the labor involved in recognizing and interpreting data, emphasizing the shared cognitive effort between humans and technology.
> Human Labor and Machine: Recognition - Recitation & Script-Training
The recitation of words in "Human Alt-Text" is not merely a performance; it represents a labor-intensive process that reflects the cognitive effort involved in data recognition. This act of recitation requires participants to engage deeply with language, demanding focus, memory retention, and the ability to articulate thoughts clearly.
In machine learning, algorithms analyze large datasets, identifying patterns and relationships to improve accuracy. Similarly, when humans recite words, they actively engage with the language, making connections and employing critical thinking to generate descriptive narratives. This engagement mirrors the iterative training processes of algorithms, where repeated exposure leads to refinement and growth.
Just as machines learn from mistakes through feedback loops, humans also refine their understanding through practice and correction. In the context of recitation, participants may stumble or misinterpret a word, prompting them to adjust their approach. This iterative process echoes how machine learning models adapt and improve based on their experiences.
bottom of page