Transformers, a type of machine learning model, have been known to work primarily on textual representations of data. These models have achieved impressive results in a wide range of applications by learning from patterns within the text. In a groundbreaking update, transformers are now being utilized to operate directly on file bytes instead of relying solely on textual data, marking a major leap forward in the field. This new approach has several key advantages, including bypassing the need for pre-processing data to transform it into a textual representation that a transformer can understand. Operating directly on bytes means that any data type, whether it is text, images or binaries, can be processed more efficiently by the model. As a consequence, this new approach has the potential to streamline the machine learning process, significantly reducing the time and resources required to train and implement transformers. The direct byte-based operation of transformers opens up a whole new frontier of possibilities in the world of AI research and application. Most significantly, it overcomes the limitations of transformers that operate solely on textual data and paves the way for innovative leaps in fields such as natural language processing, image recognition, and cybersecurity.