Generative AI: From Glitches to Blandness

Generative AI has undergone a remarkable transformation, yet this evolution comes with its drawbacks. Tim Adams references a pertinent quote from American novelist Nicholson Baker about the size and complexity of thoughts, drawing a parallel to the current output of generative AI. Initially, this technology often produced amusingly absurd suggestions—remember the infamous advice to “glue cheese to pizza”? However, the more recent iterations have become markedly less quirky, offering instead executive summaries that strip away the richness and depth of thought.

The shift to more pragmatic outputs seems welcome at first glance; however, it inadvertently results in a less engaging experience. As AI transitions from whimsical to utilitarian, it increasingly serves up watered-down versions of knowledge. The complexity and texture of human understanding are lost in the process—a concerning trend for users seeking insightful discourse.

Moreover, Adams muses on cultural references that reflect the complexity of thought and narrative. He recalls an encounter with an English teacher who preferred memorable poets over dated traditions. In a time when national identity is fluid, the call for meaningful cultural references is both an intriguing and necessary one. While he considers various poets, he eventually favors the contemporary Ben Jonson, who crafted culinary poetry that celebrated both food and fellowship.

As he reflects on the nature of storms and their naming conventions, Adams recounts an insightful tidbit about meteorologist Clement Wragge, who named tumultuous weather after politicians in an effort to articulate their unpredictable nature. This anecdote adds another layer of texture to the discussion, helping illustrate how we frame our understanding of chaos and order—principles that are mirrored in our interactions with AI.