2022.oct.21
I say this to a lot of people, I don't like deep learning, but I want to be here when it gets better. I think a more accurate phrasing is, "I don't like deep learning, but I want to be the one to make it better." I think I really could work on this problem while doing deep learning for medical imaging - the question is, am I capable enough? Even if I have a idea, do I need to be handheld to get it completed? If I'm not capable enough, can I get better? Last night at the Physics Birthday Bash (CWRU) I heard that Corbin's motto for mentoring and his work is "Do it a little less wrong next time." I would like to adopt that thinking. Maybe "Be a little bit more capable next time." I think for sure what allowed me to at least have a running start with the Wang project was my previous failures with Julia. What allowed me to also have a better start with Yong is my struggles on loss functions and inefficiencies with keeping track of runs with the Wang project. (Not to mention how much I've improved since my research years with Jesse.) From here on out, I'll try to keep in mind that I have gotten better, and I'll become a little bit more capable each time.
The next writing entry will be on scientific creativity and impact wrt age.