Navigating Mental Health and Artificial Intelligence
Over the past year, almost every industry has witnessed the growth of Artificial Intelligence (AI). Generative AI has proven especially useful, with its ability to "[create] new content, including audio, code, images, text, simulations, and videos," McKinsey & Company writes.
The most popular generative AI-powered tool is ChatGPT, a large language model (LLM) developed by OpenAI. ChatGBT produces human-like responses to a wide range of questions, making this sort of technology ideal for the healthcare industry. Although many medical professionals remain sceptical, with concerns over potential data breaches and patient privacy, it’s clear that generative AI holds great promise in the health sector. From developing personalised treatment plans to automating patient management, AI is set to transform how healthcare services support users.
With a commitment to innovation, MindSkiller® is leading the way.
MindSkiller® CONNECT will incorporate AI
While AI complements MindSkiller® in numerous capacities, an AI-powered chatbot is currently in development for the CONNECT service. Designed as a video-conferencing tool for educational support, CONNECT provides users with instant and real-time assistance from mentors, clinicians, and coaches.
MindSkiller’s AI component serves two functions. First, it can act as a triage tool, conducting a preliminary assessment of the user's needs before connecting them with a help-provider. Second, it caters to users who are reluctant to speak with a real person. Here, the availability of a smart AI chatbot will give users personalised responses drawing upon an extensive mental health knowledge-base distributed amongst over 65 educational modules and other content within the MindSkiller ecosystem.
This is particularly useful for Gen Z. As Jennifer Jones-Mitchell writes for Forbes, "Many [young people] may be more comfortable talking to an AI chatbot than a person... They've grown up in a digital age, and they've known nothing else."
Research from Biotech company Roche Products Ltd supports this assumption. Surveying more than 5,000 British adults in 2018, researchers discovered that 82% of young people (aged between 16 and 24) want the UK National Health Service to embrace "digital solutions to better manage and improve their healthcare." In fact, almost two-thirds of this demographic said they would be comfortable with a chatbot diagnosing them instead of a person. Given the increasing prominence of AI and the surge in digital healthcare as a result of COVID-19, it’s safe to assume that this percentage has increased.
MindSkiller's Gen Z consulting group is leading the platform's AI development
While some team members are researching existing chatbots and engines, others are building a specific AI model for CONNECT. Saul Galambos, the head of the Gen Z team, holds high hopes for the future - sharing that one day, MindSkiller® may be able to use AI to match users "with what they need" based on their activity on the platform.
After speaking with Luca Moses, the computing and finance ANU student developing the chatbot, it's clear that he also sees the potential of AI.
"AI is a relatively new technology, and it is becoming increasingly useful and accessible to everyone," he begins. "The opportunity to help integrate it into a medical capacity and grow the usage of AI is something that I am especially fascinated by and thoroughly enjoy."
When asked how the AI chatbot will work, he explains that by using OpenAI's Application Programming Interface (API),* he can create an application compatible with GBT-3.5-Turbo. Through this, MindSkiller’s AI chatbot will have access “to cutting-edge language and speech-to-text capabilities.”
"By using a technology called ‘embedding’, I'm able to finetune the AI model with data from the MindSkiller® website, enabling the chatbot to respond with relevant MindSkiller® related answers," he adds.
Currently, MindSkiller’s AI model relies on the glossary for information, but Luca intends to expand the range of data sources from which the chatbot can draw.
"In future, the AI may be able to also utilise information from the user's MindSkiller account, such as modules completed, and... provide a more personalised response," he goes on to say. "By restricting the AI's data sources, it does help ensure the accuracy of its responses, as it can provide personalised and MindSkiller relevant answers."
Of course, with greater access to data come concerns over patient privacy and consent. LLMs, like ChatGBT, "may not protect sensitive data (including health data) that a user provides to an application to generate a response," writes the World Health Organisation. In light of this, the MindSkiller® team is diligently working to ensure the safety and privacy of users when engaging with the chatbot.
"Currently, no user data is being stored and saved by the MindSkiller AI, allowing users to securely and safely converse with the chatbot," Luca shares. "Furthermore, 'zero-data retention' can be opted for when using OpenAI's models, which adds an extra layer of security regarding user information. This is currently being added to the MindSkiller chatbot."
See some examples of a conversation with the MindSkiller AI-powered chatbot below:
Q: What's the difference between the leading edge and tail edge of the performance wave?
A: The leading edge of the performance wave represents high anxiety, where you may experience being overfocused, overplanning, and overreacting. It is the threshold beyond which you tip over the edge and get dumped into high anxiety. On the other hand, the tail end of the performance wave represents low energy or depression. It is where you can slide off the edge after being triggered by multiple threats. In this state, you may experience low motivation, fatigue, and withdrawal.
Q: Who is Dr. Egg, and how can he help me?
A: Dr. Egg represents your imaginary Psychiatrist. His mission is to educate and train you using a psychiatrist's 'inside' knowledge, strategies, and skills. He is an expert in mental health diagnosis, research, and treatment. Dr. Egg provides jargon-free, simple explanations and strategies to help you understand and improve your mental health. He uses visual analogies to introduce mental health concepts and guide you on your journey to recovery, resilience, and mental fitness.
Check out MindSkiller® AI chatbot here.
*An API is a tool that software developers use to access data and facilitate communication between programs even if they don't share the same server.
Resources:
https://www.mckinsey.com/featured-insights/mckinsey-explainers/what-is-generative-ai https://www.pharmatimes.com/news/generation_z_expect_technology_revolution_in_the_nhs_1263533
https://www.who.int/news/item/16-05-2023-who-calls-for-safe-and-ethical-ai-for-health
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8826344/
https://www.brookings.edu/articles/risks-and-remedies-for-artificial-intelligence-in-health-care/
https://www.mckinsey.com/featured-insights/mckinsey-explainers/what-is-generative-ai
https://www.forbes.com/sites/forbescommunicationscouncil/2023/07/07/adapt-or-die-how-to-connect-with-gen-z-consumers-and-employees/?sh=44955ee52350 https://openai.com/blog/introducing-chatgpt-and-whisper-apis