JGP: July & August Update (Part 2)

Early august was spent learning how to use ChatGPT’s API, and researching current AI chat bots that deliver mental health support. I also expanded my scope to the whole “self-care” market to gain a better understanding of the success criteria in this market. 

After comprehensive research, I’ve identified fundamental flaws in currently available mental health AI Chatbots, which I believe I can capitalise on and improve overall service quality. Thus, I further invested myself into studying ChatGPT’s API and other related areas so I can experiment with creating a Chatbot that offers mental health support. 

In the meantime, I used university resources to self-study Cognitive Behavioural Therapy (CBT). Thanks to my foundations in Psychology and Language Sciences, I am able quickly learn the theory and processes behind CBT, and start integrating CBT treatment approaches and exercises into my Chatbot. 

As of today, I have built “Li” (named after my school teacher who greatly influenced the trajectory of my life), a generative Chatbot that can offer some mental health advice based loosely on CBT. A major achievement (for me, perhaps not to others) was when I figured out how to give Li contextual knowledge, since up till that point Li was unable to refer to the conversation and make contextually relevant responses. Giving Li contextual understanding greatly improves conversation quality and coherency.

However, Li is far from finished and in the coming months I will be looking to create more solid CBT-protocols within Li, and develop an MVP to gage market interest in using my Chatbot as a supplement to traditional therapy. 

On a side note, a videographer expressed interest in turning my project into a documentary. She and her team will be following me for a whole year to document the progression of my project and release monthly/bi-monthly episodes. I recently got the “OK” from my program director so we’ll see what goes on from here.