In the latest round of ChatGPT updates, the focus is no longer just on “answering better,” but on making conversations feel more like working with a real assistant: you can speak more smoothly, it can remember your preferences, and privacy controls are put back in your hands. Add file analysis and direct cloud-drive connections, and efficiency in handling materials improves noticeably. Below, I break down these new changes by feature—how to use them and what their limits are.
Advanced Voice Mode: More Natural Conversational Rhythm and Expression
ChatGPT’s Advanced Voice mode is being rolled out in batches, with key improvements focused on response speed, stability, and voice realism. In practice, you can treat it like a spoken conversation where you “talk and follow up as you go,” without having to type your question into a long block of text first.
If you haven’t seen the voice entry point yet, it’s usually because the feature hasn’t been rolled out to your account—not because your settings are wrong. It’s recommended to use it on a stable network and to clearly state your goal in the conversation, such as “give the conclusion first, then explain,” so ChatGPT’s spoken answers will be more concise.
Memory Feature Is Here: Let ChatGPT Remember Preferences—And Clear Them Anytime
The memory feature is now available to more ChatGPT users. It can keep your repeatedly mentioned preferences “effective over the long term,” such as your commonly used language, writing tone, work background, and so on. More importantly, ChatGPT will prompt you when it updates memory, so you know exactly what it has remembered.
The control options are also more granular: you can turn off ChatGPT’s memory, delete a specific memory item, or use a temporary chat to avoid being recorded. For people who care about privacy, this is more controllable than “remember everything by default.”
