I'm a quantum physicist turned entrepreneur. I enjoy thinking about my old research still, and am not motivated by the peer-reviewed publication games of academia, instead, I'm obsessed with how AI can super-boost physics, and science-general, research. I used AI (claude Opus 4.8) to push the research envelope on my original publication (high-dimensional QKD using time-energy entanglement), and was very pleased with the results. I built a template to let research be open-sourced like software, easily shareable, easy to correct and collaborate like code. This could boost the speed of research, compared to the decades old structure of canonical peer-review. With the template I am sharing, you can share research on github with a referenceable DOI, and timestamps, for provable priority. This is the direction that my research is heading towards: AI discovering possible research, doing the research, publishing the research, and repeating -- with critical human oversight, which should improve the process over time. In this dossier structure, the research results are presented in both the easy-to-read (beginner) format as well as the canonical expert format. Humans can learn as they read, and critique the results, just like bugs in code.New QKD Protocol: https://m4gr4th34.github.io/dossier-001/
Template: https://github.com/m4gr4th34/open-dossier-template