This chapter carries out an analytical condensation of the consciousness issue and corrects the lack of authenticity in descriptions of human/machine relations. As a consequence, KB/Chat sessions are initiated to clarify how close we are to the future ‘machines with self- consciousness’. The conversations move across the major topics from consciousness to abilities to ‘analyze and understand language’ to machine learning. KB follows Chat into the machine room and listens to Chat’s I-strong statements. Against this background, an analytical strategic clarification of Chat’s self-understanding and understanding of AI’s characteristic human-machine/relations and tasks is carried out. Here, the conflict between humanized abilities and technical-machine actions is made visible. In the analysis, the humanized abilities are framed by Chat’s extensive presentations of ‘a broad range of cognitive skills’, which have ‘to analyze, understand language’ and ‘to learn’ as absolute basic skills. Chat’s I-strong presentations function here as a statement that plays together with the IT expertise’s description of the programmed abilities of AI/ChatGPT. That means that Chat’s statement ‘I can generate language’ is credible and that IT expertise is behind it. On the other hand, when the chapter completes the analysis, the conflict is seen: there is no professional or expert knowledge behind the statement “I can learn from data…”. The statement, the sentence and the entire statement then suffer both a loss of reference and a loss of meaning. At the same time, Chat’s assertion of his own abilities raises a question of credibility and responsibility. The question of responsibility is made visible in the machine room, but then it appears in the future as a question of doubt in the analysis. There is a constant doubt about who takes responsibility for Chat’s mechanical actions and task solutions.