The future of experimentation and innovation in language models poses a question of whether to focus on enlarging data sets and models or on developing smaller models with greater efficiency. However, a balance between the two approaches seems crucial. The enhancement in language models like Claude three has enabled generating compelling content with substantial context, blurring the lines between human and AI-generated content. This advancement showcases the potential of large models in assisting with content creation and authoring, indicating the continuing significance of scaling up models for such creative endeavours.