ChatGPT's Branching Update Sparks Important Reminder: AI Chatbots Have Limits

ChatGPT's New Branching Feature Reminds Us AI Chatbots Aren't People On Thursday, OpenAI announced a significant update to its popular chatbot platform, ChatGPT, allowing users to branch conversations into multiple parallel threads. This feature, years in the making, serves as a timely reminder that AI chatbots are not people with fixed viewpoints but rather malleable tools that can be rewound and redirected. The new branching capability is now available for all logged-in web users, who can access it by hovering over any message in a ChatGPT conversation, clicking "More actions," and selecting "Branch in new chat." This creates a new conversation thread that includes all the conversation history up to that specific point while preserving the original conversation intact. The feature is designed to enable users to explore different ideas and scenarios within the same conversation, much like creating multiple drafts of a document. According to OpenAI, the branching feature was developed in response to user requests for greater flexibility and creativity in their conversations with ChatGPT. "We're excited to provide our users with more ways to engage with our chatbot," said an OpenAI spokesperson. "This new feature will enable them to explore different ideas and scenarios within the same conversation, making it even more useful for tasks like brainstorming and research." The branching feature has significant implications for various industries, including marketing, customer service, and education. For instance, a marketing team can now create separate branches to test different ad copy approaches or tone, while preserving the original conversation history. Background and Context ChatGPT is an AI chatbot developed by OpenAI that uses natural language processing (NLP) to understand and respond to user queries. Since its launch in November 2022, ChatGPT has gained widespread popularity for its ability to engage in conversations on a wide range of topics. However, the platform's limitations have also been highlighted, including its tendency to stick to pre-programmed responses rather than exploring new ideas. The branching feature is a significant step forward in addressing these limitations and enabling users to harness the full potential of AI chatbots like ChatGPT. Additional Perspectives Dr. Joanna Bryson, a leading expert on AI ethics, welcomed the development as a positive step towards greater flexibility and creativity in human-AI interactions. "This new feature is a great reminder that AI systems are not people with fixed viewpoints but rather tools that can be rewound and redirected," she said. However, some experts have raised concerns about the potential for branching conversations to create confusion or miscommunication between users and chatbots. "While this feature may seem useful on the surface, it also raises questions about accountability and responsibility in AI-mediated interactions," noted Dr. Bryson. Current Status and Next Developments The new branching feature is now available for all logged-in web users of ChatGPT. OpenAI has announced plans to continue developing and refining the feature based on user feedback and testing. As the use of AI chatbots continues to grow, this development serves as a timely reminder that these systems are not people but rather tools that can be shaped and directed by human creativity and ingenuity. *Reporting by Arstechnica.*