Listen: Your 'Gmail' can write code.
Gemini in Gmail drafted Python code to fine-tune a DeBERTa model for sentiment classification, complete with data loading, training loops, and evaluation.
Transcript
The Gmail drafting tool is powered by a robust language model, and it is capable of doing almost anything a standard chat assistant can do. To test its limits, you can ask it to write complex code right inside your email window. In one instance, the drafter successfully generated a fully functional Python script to train a machine learning model. It wrote code to load data, set up a sentiment classification system, and even handle logging and progress tracking. It then packaged all of this into a ready-to-send email draft. Its capabilities go beyond writing and coding. You can even ask the Gmail drafter to design a new logo. While you might decide to stick with your current branding, the experiment proves a larger point. The tools built directly into our everyday office applications have evolved far beyond simple text completion. They are now versatile, general-purpose assistants ready to tackle highly technical and creative tasks.
