Recently, ChatGPT has made clear progress in two areas: “understanding you better” and “being more useful.” On one hand, it has filled in the gaps around memory and privacy controls; on the other, it has made files, voice, and the desktop experience smoother to use. Below, organized by usage scenarios, is a quick guide to what problems these new features can solve.
ChatGPT Memory: what it remembers and how you can manage it
ChatGPT’s memory feature records information that’s useful to you over the long term at appropriate times—such as your preferred writing style, commonly used language, or fixed formatting requirements—so future conversations cost less to explain. It doesn’t “permanently save” all chat content; instead, it extracts reusable preferences in a more controllable way.
More importantly, you’re in control: when ChatGPT updates its memory, it will notify you. You can review and delete specific memories, or turn the memory feature off entirely. For people who don’t want personalization to influence them, these toggles make ChatGPT feel more like an “adjustable” tool rather than a black box.
Direct cloud-drive uploads: feed Google Drive and OneDrive files to ChatGPT
In data analysis scenarios, ChatGPT supports direct uploads from Google Drive and Microsoft OneDrive, eliminating the need to repeatedly download and then re-upload files. For frequently updated documents like spreadsheets, reports, and project materials, this change noticeably reduces friction.
After uploading, ChatGPT can summarize, compare, and explain the data, and can also generate visualizations such as charts, making it easy to use directly in presentation materials. When asking, it helps to be specific about the goal—for example, “Create a monthly trend chart and mark the anomalies.”
