Show HN: MoodMinder – Swift Anger Regulation for Better Emotional Well-Being (moodminder-34552.bubbleapps.io)

18 points by marans ↗ HN
Hey Hacker News community! We're excited to showcase MoodMinder, a mental health app MVP that empowers individuals to swiftly regulate anger and enhance emotional well-being. MoodMinder was born out of a desire to provide quick anger regulation solutions for busy individuals.

Unique Features:

Rapid Mood Identification: Identify anger triggers and tension levels swiftly.

Instant Personalized meditations: Receive tailored meditations for immediate anger control.

Cognitive Reappraisal: Shift perspectives to defuse triggers in real-time.

Quick Interactive Games: Engage in games designed for anger regulation in just minutes.

As an MVP, we're seeking insights from the Hacker News community to shape our app's development. Your feedback is pivotal. Thank you for being part of our journey!

30 comments

[ 1.6 ms ] story [ 79.3 ms ] thread
Interesting project! How do you model emotions and what is the source of data in MoodMinder?
We appreciate your curiosity about our project! It is really precious for us and your engagement fuels our progress. To shed light on your questions, let us delve into our approach:

At the core of MoodMinder is a conceptual cognitive model, meticulously crafted to regulate emotions. Emotions are a central pillar of this model, as they significantly influence our perception and interactions with the world. To test and validate our model, we've integrated the power of GPT into our framework. We utilize GPT's remarkable ability to understand and generate human-like text and a game to customize our training process. This synergy empowers MoodMinder to fine-tune its emotional intelligence. As we continue to enhance MoodMinder's capabilities, the fine-tuning of GPT remains a pivotal step. This process ensures that the model aligns even more closely with the nuances of human emotions, creating a symbiotic relationship between AI and human emotion understanding. Thank you for your interest in our journey! We're excited to have you accompany us as we explore the fascinating landscape of emotions through technology.

Are you using GPT to author or edit your responses?
We utilize ChatGPT to enhance general responses, making them personalized and efficient for each user's unique inquiry.
Small piece of feedback:

GPT can be useful for editing purposes, esp. for people writing in their second language or trying to find a more concise way of expressing themselves. (I'd often throw a phrase at GPT to see if it feels natural/idiomatic, instead of bothering my better half - an editor)

Having said that, if your goal was to make the responses here (on HN) more personalised, you achieved the opposite. Perhaps this is a sign of my personal paranoia, but I'm still not sure if I'm talking to a person or a bot. I don't mean to sound harsh, just thought it might be useful for you to know.

(FYI, I meant the responses here on HN. )

I genuinely appreciate your feedback and want to assure you that I'm indeed a real person responding to all comments. I'm curious to understand more about why you feel that way – could you kindly share more insights? Your input is highly valuable in shaping our approach.
Because you respond to things in a monotone style with robotic/chat-bot-like cheeriness.

Your comments lack empathy and you aren't "reading the room" at all.

This is totally off-brand and sleazy if you're selling mental health advice. Particularly the fact that your responses don't demonstrate natural emotional variation, when you are literally selling a tool to help with emotion modulation.

I appreciate your feedback and apologize for any unintended tone. English isn't my first language, and my focus on accuracy might lead to an emotionless perception. Your point about empathy is valid, and I try to improve my responses.
I think if you used ChatGPT-4 to write in your native language and ask it to translate it to English you might get good results. There are some minor tweaks I like to use when translating with ChatGPT:

From another language: “Translate this to English. Provide any clues you can get from idiosyncrasies in spelling, grammar, or word choice which might indicate which specific region or sub-culture the writer might be from and use this to provide notes on deeper or alternate meanings specific to that region or sub-culture. Note any potential ambiguities or uncertainties in translation and provide multiple options where appropriate. Use context clues to explain the tone, tenor, and emotion of the writing/writer, and what that commonly means in this culture vs. American culture.”

To another language: “Translate this to $LANGUAGE. Please make it sound natural to a native speaker, use local idioms or phrases wherever possible. Rather than strictly transliterating it, focus on conveying the specific tone of the original, but note where there may be a conflict between cultures and inform me if there’s a risk of being misinterpreted.”

on the third or fourth screen in meditation is spelt 'mediation'.

hth

...this could make some people VERY angry :)
Thank you for bringing this to our attention. Fixed
I love that you want to help but is this, like, reviewed by mental health professionals?
Absolutely, your concern is valid! Our product has been crafted by a team of cognitive psychologists to ensure its foundation is rooted in expert knowledge and insights. Your well-being is at the heart of what we do.
There needs to be detail about who you are on the site. Credentials and/or what research group or university affiliation.

Also needs privacy statement about how data is handled. And with regards to what is sent to open AI.

I wouldn’t use this at all. There are 3+ entities (bubble, bubbles infra, openai, yourself) with access to super sensitive information about me, people I know etc.

Appreciate your input! We're in our early stages and value your concerns. Bubble is for development only; we don't share data with OpenAI or collect user info. Rest assured, no personal data is shared or stored on our servers. Stay tuned as we enhance our MVP and provide full credentials and affiliations for transparency.
> we don't share data with OpenAI

Yes, you do. You state on the front page that this is built on ChatGPT.

We appreciate your concern. To clarify, while we do utilize ChatGPT for our project, we don't share personal data with OpenAI. Your privacy and data security are paramount to us.
Most of the "data" people will be entering into this application is deeply, deeply personal, even if not in the way you probably mean (PII).
We've taken note of your concern and made email input optional. As we're in our initial development phase, your suggestions to enhance this aspect are greatly appreciated. If you have further insights to ensure a safe experience, we're all ears.
I have an inherent and probably incurable distrust of applications that send unknown content to OpenAI or other LLM services. I am unlikely to ever use products that do this.
Your concerns about data sensitivity are understood, and we're committed to continuous improvement in our updates and future versions. Your viewpoint is respected, and while products vary in fit, we're open to your suggestions. Your feedback is highly valued. Have a wonderful day!
(comment deleted)
This response sounds very insincere.

Edit: I liked the explanation provided by someone else in this thread that you have already replied to, so you can disregard this comment.

This response style is deeply unsettling to me. It fails to convey thoughtfulness or sincerity, at least in the cultures that I was raised in.

I have major concerns about the scope of what is fed to ChatGPT vs. GPT-4 API. I am also concerned by knee-jerk reactions here which take minimum action that seems to continue to focus around formal PII without an introspective review of customers real concerns, which is the totality of personal info likely to be shared with your service and how that’s handled and treated.

This isn’t written for you — I worry if you respond at all you’ll send a generic imitation of corpo-speak “We respect your views and we’re working on it!”. This is written to others, to provide a jumping off point to discuss what makes this style of communication stand out so much, and what signals it provides customers vs. what signals it provides investors.

This is great! A novel and interesting use for AI (I played the game). However, I have a request. Please add a privacy policy for collecting emails at the end. Thank you!
Thank you for your positive feedback and for trying out the game! We've addressed your request. We will add a privacy policy regarding email collection in a few minutes.