ChatGPT’s new branching feature is a good reminder that AI chatbots aren’t people (arstechnica.com)

OpenAI has introduced a new branching feature for ChatGPT, allowing users to create multiple parallel conversation threads from any point in a chat. By selecting "Branch in new chat," users can duplicate the conversation history up to a chosen message and explore alternative directions without losing the original dialogue. This addresses a key limitation in previous linear chat interfaces, where users had to either overwrite prompts or start fresh to try different approaches. This feature is significant for the AI/ML community as it transforms how users interact with large language models, enabling more flexible, efficient experimentation. It supports complex workflows like brainstorming, structured learning, and project analysis—tasks that benefit from exploring multiple "what-if" scenarios in parallel. The branching functionality parallels version control systems like Git, emphasizing the evolving view of AI chats as editable, evolving documents rather than fixed conversations. Technically, the feature reduces cognitive load by removing the need to manually copy or compare responses across conversations, as highlighted by a 2024 study from Tsinghua University and Beijing Institute of Technology. By making interaction with LLMs more dynamic and multi-threaded, OpenAI advances user control and creativity, expanding practical applications of AI chatbots beyond simple linear dialogues.