Show HN: Orchestrate frontier AIs on the subscriptions you already pay for (github.com)

A new project called PolyHelper allows users to efficiently orchestrate multiple AI models to answer questions using their existing subscriptions, without requiring any installations or additional costs. Users can query various AI services—whether free or paid—collect their responses, and receive a consolidated document that highlights points of agreement and disagreement among the models. This method is designed to help users obtain nuanced insights and mitigate the risks of relying on any single AI’s potentially false confidence. The significance of PolyHelper lies in its ability to enhance the decision-making process by providing a clearer picture of different AI perspectives. By leveraging the current models like ChatGPT and Claude, PolyHelper creates a “council” effect, where multiple AIs weigh in on a question, presenting not just a summary but also revealing discrepancies in their responses. This capability is particularly useful for critical decisions, such as selecting a product or service, where understanding the consensus and areas of contention among AI outputs is crucial for accuracy. Furthermore, the approach facilitates engagement with AI in a more interactive way, empowering users to generate detailed insights from existing tools they already use.