GPT‑4.5: The Silent Revolution in Artificial Writing

The evolution of language models has been rapid since the publication of the paper Attention is All You Need, which laid the foundation for the Transformer architecture. In the following years, with the advent of large-scale models and the emergence of ChatGPT, the revolution accelerated: systems were refined to solve mathematical and programming problems and offered increasingly polished and natural writing. Multimodal models capable of generating videos and images also emerged, along with internet-integrated searches and conversational developments that broke records across various benchmarks, especially in logical and mathematical domains.

However, until the arrival of GPT‑4.5, no model had managed to achieve truly remarkable text composition. Few models can produce sophisticated texts, using similes, analogies, and literary devices that reflect meticulous craftsmanship, inspiration, talent, and experience. An average human—unless they have read and practiced extensively—struggles to reach such a high level of writing excellence. For the first time, GPT‑4.5 has demonstrated that a machine can write at a superlative level, surpassing the expectations of those who believed writing proficiency was an exclusively human domain.

On Reddit forums, opinions on GPT‑4.5 are divided. Many users criticize that, despite the marketing hype, its improvements are marginal compared to previous versions, particularly in coding and reasoning tasks. However, these critics often compare trivial results, such as the ability to count the number of “r’s” in the word strawberry. This reveals a lack of understanding regarding the true breakthrough that GPT‑4.5 represents in the field of writing.

The model not only excels at solving logical or mathematical problems but also exhibits a remarkable ability to generate texts akin to those crafted by experienced writers. This marks a paradigm shift in how we perceive artificial intelligence applied to writing. Until now, AI was primarily used to generate generic texts—where it was evident that the content was machine-produced. GPT‑4.5, however, opens the door to texts so refined that distinguishing between human and AI-written content becomes nearly impossible.

The impact of this technology could be as transformative as any cutting-edge innovation. While the percentage of people who program is relatively small, the vast majority of the population interacts or creates through the written word. The ability to write exceptionally well is a deeply human skill, and the fact that a machine can achieve such a level marks a milestone that could radically change various sectors, from communication and education to creativity as a whole.

In summary, while some perceive GPT‑4.5 as offering only marginal improvements in technical aspects, its ability to generate high-quality texts—with sophisticated literary techniques and impeccable writing—represents a revolutionary advancement in artificial intelligence. This model, still undervalued by those who focus solely on logical benchmarks, invites us to rethink AI’s potential in literary creation and to imagine a future where the boundary between human-written and machine-generated text becomes practically indistinguishable.