PsyDactic - Child and Adolescent Psychiatry Board Study Edition
Using the American Board of Psychiatry and Neurology content outline for the Child and Adolescent Psychiatry board exam, starting with the most high yield, Dr. O'Leary has created this podcast for anyone interested in CAPS and also to help him study for the boards. Enjoy!
Let Dr. O'Leary know what you think by going to https://psydactic.com/ and filling out the form there.
PsyDactic - Child and Adolescent Psychiatry Board Study Edition
001 - Trailer and Introduction of PsyDactic - CAPS Edition
Dr. O'Leary introduces PsyDactic - CAPS Edition, explains the goal of this podcast, and how it was produced.
This podcast is intended as a study aid for the American Board of Psychiatry and Neurology (ABPN) Child and Adolescent Psychiatry board exam.
Feedback can be submitted via a form at https://psydactic.com.
This is not medical advice. Please see a licensed physician for any personal questions regarding your own or our child's health.
00 - Episode 1 - The making of PsyDactic - Child and Adolescent Board Study
Welcome to PsyDactic - Child and Adolescent Psychiatry Board Study Edition. Let’s call it CAPS Edition for short. I am Dr. O’Leary and as I speak these words, I am a child and adolescent psychiatry fellow in the national capital region. I am also the host of the podcast PsyDactic, which I started during my second year in residency. This podcast is part of the expanding PsyDactic family. This project is my attempt to both study for my board exams and also create something that I could then listen to myself on my way to, well, anywhere. I hope to be able to publish at least one episode per week on average covering all of the major topics on the American Board of Psychiatry and Neurology content outline for the CAPS board exam, starting with the most high yield.
Before you commit yourself to this podcast, I want to consent you to a few things. The first is that I do not have an editorial staff, so what you get here is pure me and my resources. Everything here, even if it is masquerading as a fact, is really just my packaged version of that fact. It is the version that I wanted you to hear, the version I believe. I claim full responsibility for these opinions and they should not be mistaken for the opinions of anyone else or any institution, whether they currently employ me, or just tolerate me enough to let me visit, or to hang out in their parking lot.
The second major thing is that this project represents my first major attempt at using AI to help me create content. I have dabbled before, but what you are getting here is a tad bit more than a dabble. My basic process is this:
I generally approach creating content by asking an AI to generate a general outline of the concepts that I need to cover. I then go through a process of checking and editing the content for accuracy and completeness. I then might ask for the content to be generated in different ways. I have been learning more and more how to construct prompts that increase the chance that I will get more complete and accurate answers. I guess you could say that the AI has been teaching me how to make it better.
Before finishing and recording any episode, I do a last comb through, just to try to catch any mistakes. There is no way that I am going to catch them all, so if you want to let me know what I have missed, please go to PsyDactic.Com (https://psydactic.com/) and fill out the form there. You can be part of making this better.
I also want to make a statement about the fraught ethical and practical implications of using AI for this project. There are many. AI large language models are only as good as their architecture, programing and training data. Much of that data has implicit biases, racial, ethnic, gendered, cultural, political, national, etc. Also, and probably more problematic from a practical point of view, by producing content that AI contributed to, I may be helping to train future AI models on AI generated content, which can contribute to a few more problems. One of these is what I think of as amplification. AI learning from AI can amplify the errors and biases within AI models. Imagine sending a picture through a filter and then another filter and then another filter. What comes out at the end could look very different. It might have thick lines around whatever parts of the image that the AI thought was relevant, amplifying their importance in the composition. Also, training AI on AI generated content can cause model collapse. The metaphor I use for this is feedback into a microphone. What ends up being produced if this process continues is nonsensical, cacophonous gobbledygook.
That being said, I implore you, as the listener, to be highly critical and suspicious of this content. It would serve you well to not be compelled by the illusion of knowledge and quit asking questions because it appears that everything has been answered. I will strive to raise doubts as I discuss each topic so that I don’t give the impression that this content is, by itself, authoritative. It is intended to be a tool to help CAPS fellows get through their boards with substantially less stress.
If you think your friends or colleagues will like this, I am on all of the major streaming platforms. You can also go to PsyDactic.buzzsprout.com for my original podcast or and my RSS feed.
Cheerio, toodle-ooo, and happy listening. I am Dr. O’Leary, and this has been the first episode of PsyDactic - CAPS edition.