This is a pilot project exploring how generative AI can be utilised in Radical Technology Inquirer (RTI) method to make RTI less labor-intensive and to learn how Retrieval-Augmented-Generation (RAG) is best used in technology foresight. RTI is a technology foresight method that has been first developed in 2013 and improved in 2018 as a project commissioned by the Committee for the Future. RTI follows the development of 100 radical technologies and their impacts on 20 different societal functions (logistics, health, nutrition, production of experiences etc...). AI-TF-HumRob explores the future impacts of just one radical technology that is currently attracting considerable investments in e.g. USA and China - humanoid robots. The humanoid robot development was chosen as the pilot case, because several independent development projects already show considerable humanoid robot capabilities. Should the investments continue, the capabilities are likely to improve while the cost of human-like assistance will decrease. This dynamic is further accelerated, if the AI-driven robots are made capable for autonomous reinforcement learning and if they they upload any learned skill to all robots in the same network, spreading them quickly across the globe. In the long run, this dynamic will have huge impacts. AI-TF-HumRob explores what these impacts could be across the different functions of society. AI-TF-HumRob pilots the utilisation of generative AI (Chat GPT-o1) as a supporting tool in the analysis, production and cross-examination of texts that are required to perform RTI analysis. RTI is a suitable context for exploring the potential of generative AI in technology foresight, because in the RTI method the empirical material is text-based (technology news) and intermediate and final results are generated according to strict rules and presented in a very structured textual form. The human analyst has created the initial descriptions of present humanoid robot developments based on web searches and refined the 20 descriptions of societal functions (originals to be found in the 2018 RTI report). This text-intensive basis of RTI allows benefiting from the best feature of RAG, which is to automatically process large quantities of qualitative data. The strict methodological rules provide a good basis for developing prompts for looking for potential future impacts. This is done by systematically cross-tabulating different datasets with prompts that returns results that are credible and acceptable for the human analyst that knows the original material (descriptions of technological development and societal functions). The intermediate and final prompts and AI-responses leave a transparent track record of the foresight process.