OpenAI Tweaks GPT-5.5 Instant, Adds 5-Reply Chat Navigation and Mobile Gesture

5 articles · Updated · 9to5Mac · Jun 1

Friday’s update changed GPT-5.5 Instant—the default ChatGPT model for most prompts—to make replies easier to read, more natural and better paced, with fewer overly long or bullet-heavy answers.
OpenAI also added a mobile long-press on the send button that opens an effort picker before submission, letting users choose Instant, Thinking or Extended depending on their subscription tier.
On the web app, chats with 5 or more responses now generate a table of contents that appears from a right-side icon and lets users jump between topics in long conversations.
The changes extend OpenAI’s push since GPT-5.5 Instant launched on May 5 to streamline ChatGPT’s default experience while relying more on automatic model selection.

ChatGPT's new auto-outline seems minor. Is it a convenience, or a sign of how AI will fundamentally restructure human thought?

ChatGPT now sounds more human. Is this improved user experience masking deeper risks like bias and the erasure of unique dialects?

OpenAI's new feature lets you control AI "effort." Does this truly empower users or just create a powerful illusion of control?