As AI tools become more embedded into everyday life, discussion has increasingly shifted from understanding the systems themselves to simply managing their presence and consequences in our lives, institutions, and governments. But AI is not abstract magic. At Creative Applications, artists weigh in on machine “thinking”.
Popular LLMs are designed systems, built through human choices, technical architectures, policies, and assumptions. As things continue to shift towards more AI adoption, I often turn to two widely discussed papers to help ground the conversations surrounding it, and remind me of a few key moments in time that contextualize AI discourse. The first is Google’s 2017 paper Attention Is All You Need, which introduced transformer architecture, which is the foundation of most contemporary large language models. The second is Apple’s 2025 paper The Illusion of Thinking, which questions whether current AI systems can genuinely reason or if they only produce convincing or technically correct outputs. Together, these papers remind us that AI systems, and the solutions proposed around them, are all human-made. Though Apple and Google are among the tech giants of today, their teams of technologists and researchers are few and specialized. They prototype, run tests, and show their findings similar to university academics and public artists. Looking closely at their work as work can helps us remember that we do not have to be passive observers of technological change and that these findings can be questioned, challenged, and, when needed, called to account.
Pulling themes from Apple’s paper, Creative Application published The Illusion of Thinking, a book platforming artists’ thoughts on AI. The book is a play on Apple’s earlier paper, and it aims to challenge and extend Apple’s querying of machine ‘thinking’ through an interdisciplinary mix of artistic and critical research. They examine the overlooked implications of artificial intelligence and its assumed ability to think and reason. While AI often suggests genuine cognition, it primarily mirrors human ideas through recall and imitation, raising fundamental questions about the nature of understanding itself.
The book includes contributions by Flavia Dzodan, Nora O’ Murchú, and Florian Weigl, as well as interviews with artists and researchers such as Constant Dullaart, Rafael Lozano-Hemmer, Femke Herregraven, and Dirk Paesmans (ID.ACCO). Their work addresses topics including algorithmic decision-making, surveillance, digital infrastructures, and the role of images and interfaces in organizing perception and power.
Leave a Reply