Quick Buzz Feed
Leading voices have described generative AI as an “existential risk” that could spell the end of humanity. Not so fast.
Historically, new technologies from the printing press to smartphones have evoked fears of societal destabilization. However, the latest wave of generative artificial intelligence (GenAI), including platforms like ChatGPT and Claude, has generated unprecedented anxiety. Leading figures, and even religious leaders like Pope Leo XIV, have voiced concerns, labeling GenAI an "existential risk." The authors propose that the field of design offers a less anxious, more informed perspective on AI's actual capabilities and limitations, drawing from a decade of practical experimentation with generative systems.
Generative AI, specifically systems based on GPTs and large language models (LLMs), is characterized as a sophisticated imitation machine rather than a truly inventive force. Its roots trace back to digital artists experimenting with generative adversarial networks (GANs) since 2017. The design community quickly grasped that GenAI creates "new things out of old things" through assimilation and transformation, not original creation. For instance, in architecture, AI can re-envision existing buildings in various visual styles by identifying and applying learned patterns, without possessing a human-like understanding of architectural principles.
Tools like ChatGPT and Claude function primarily as "translation machines," capable of converting content between human languages (e.g., German to Greek), refining writing (poor draft to well-structured preprint), or even translating human intent into programming code ("vibe coding"). However, a significant and seemingly inherent limitation is their tendency to "hallucinate" when prompted to generate detailed new information from sparse inputs, underscoring that their output is based on probabilistic pattern-matching rather than genuine understanding or invention.
Imitation is a fundamental aspect of intelligence and learning, playing a vital role in human development (language acquisition, craft skills) and serving as a benchmark for machine intelligence, as seen in Alan Turing's "imitation game." Despite its importance, the article emphasizes that imitation is distinct from true invention. In design, GenAI is used for initial exploration of possibilities, but the critical step of moving beyond existing data patterns and genuine invention remains a human domain. This leads leading AI researchers to describe GenAI as a "stochastic parrot," a technology that merely repeats the most probable sequences from its training data.
While GenAI will undoubtedly have substantial economic consequences by automating imitation-centric tasks such as translation, text editing, and code debugging, the design sector has not experienced a significant transformation in employment over the last decade. Core creative inputs, along with the final evaluation and judgmental aspects of design, continue to be exclusively human responsibilities, suggesting that AI has not supplanted these critical human roles.
The authors conclude that instead of succumbing to extreme views of AI as either utopian or dystopian, we should view it as a challenge to understand and reaffirm what is uniquely human. Despite its evolving capabilities and the potential for new AI systems to emerge from commoditized coding, AI does not replace human ingenuity. The most profound human ability is not imitation but invention—the unique capacity to conceptualize and bring forth what has never been seen before. This perspective suggests a future where AI augments human creativity without diminishing the essential human role in true innovation.