In this talk, David Dalrymple discusses his perspective on achieving uploading through a 10-year AI timeline. He presents a concrete plan that includes approaches that he believes do not work, such as neural dust with ultrasound alone and electron microscopy with lipid staining. He suggests that expansion microscopy, coupled with immunofluorescence, might be a viable approach for understanding synapse receptors but acknowledges challenges in visualizing neurons. Dalrymple proposes incorporating additional tools like light sheet fluorescence and interventional experiments on organoids and human brain slices to address the translation bottleneck. He also mentions the use of AI to fill in parameters based on data obtained from destructive imaging. The scalability of the project relies on automated sample slicing and distribution, as well as parallel imaging platforms. Dalrymple emphasizes the importance of validating findings using human brain slices and organoids with human genomes. He acknowledges the challenges and feasibility of the proposed approach but remains hopeful. Lastly, Dalrymple discusses the challenges in transferring data between the dense connectivity of the entire brain and the potential use of adware and mapping algorithms for optimization. He also mentions the role of lipids in determining synapse connections and the difficulty in tracing axons.