Abstract: In this response to Downes' very substantial article I can touch on only a few points. I have chosen to pass over the technical comments which might be made (e.g., concerns about reliability and availability with the proposed system of distributed resource descriptions, or praise for his honest criticism of the intelligent agent paradigm) and focus on pedagogical issues raised by his description of the possible future of educational technology. I will particularly focus on my own views of the role of humans in the automated, intelligent, semantic web future of educational technology.