Skip to content

Instantly share code, notes, and snippets.

@goatandsheep

goatandsheep/ai-mom.md

Last active March 29, 2026 05:17
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save goatandsheep/9f037a64bf879abca41c9e2a255c5f7b to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save goatandsheep/9f037a64bf879abca41c9e2a255c5f7b to your computer and use it in GitHub Desktop.
Download ZIP
Raw
ai-mom.md

Meeting of Minds - AI 2.0

2026-03-29

how AI will impact life for next 5 years

how AI will impact life for next 5 years

Resources:

  • automation consultant: https://clockout.ca
  • upcoming AI Guidebolt.com
  • privacy-first model: lumo.proton.me

Speech:

  • I'm Kemal, I am here instead of my parents Mini and Nadeem

  • I've been asked to give a quick introduction to AI

  • My friend and I are currently building Guidebolt, a Canada-first competitor to ChatGPT that is completely accurate, but we'll talk about that after the end of the meeting.

  • I'm a software engineer from McMaster University and have started a few software businesses

  • This is my second Mom meeting, the last one I spoke on cybersecurity. I hope we can help give some quick, practical advice before the start of the meeting. So let's get started

  • firstly I want to briefly explain what AI is:

  • it is a software that is meant to work like a human brain to help with calculating things, looking things up, ideas, as well as for photo and text generation

  • One interesting thing about this meeting is the name AI 2.0 because we are in the age of AI v2 not just the second meeting of the minds meeting about AI. AI used to refer to machine learning which was pattern recognition for a narrow use case (think speech to text). Now with AI v2, we have general purpose AI that is language-based. I held onto this magazine from 2017 as a reminder of AI v1, what we expected v2 to look like and the problems we faced with v1 such as racial bias. Some of them still exist.

  • creating these "brains" or AI models is very data- and energy- intensive so only a small handful of companies are creating them, such as Google's Gemini, OpenAI's ChatGPT, and Anthropic's Claude. All of them have apps you can download onto your mobile devices.

  • Some companies have integrated these brains into their products, called AI wrappers, such as Microsoft Copilot which uses ChatGPT under the hood. It's available inside Microsoft Office products to help you generate presentations, formulas and calculations in Excel for your data, and summarizing or rephrasing unclear text in Word or Outlook. All these features are still available within ChatGPT, but the interface is different and the data is already on your computer. There are a lot of AI wrapper tools out there for email, videos, research, all sorts of things and because this is a hot time for AI, there are new ones that come and go every month. This makes it hard to find what you're looking for but my suggestion is to be very pragmatic and don't worry about what is best for everyone else-just what works for your purposes today. Or the stable thing is to just use the model providers' Chat apps

  • The big model providers more or less will stay the same and the differences for most general use cases is minimal.

  • So here's an example on how I've used AI- I'm planning my best friend's wedding, so we generated text for emails, invitation text, posters and graphics. We gave ChatGPT the couple's photos and it made stylized cute photos of them

  • Other use cases could be generating recipes based on what's in your fridge or the best time to plant certain seeds based on your soil and location

  • The problem is that it can make mistakes so it's still important you use your judgement

  • One other problem is privacy. Whatever data you put into the models, they use for training and thus can be and have been extracted. There are some European models like Lumo that do not collect your data. Please be very careful with what data you provide AI

  • One thing I wanted talk about is threads. When you have a conversation about a topic with an AI Chatbot, it's important to keep as much similar information within the same discussion. The AI will use everything you give it from the past to help it with future questions.

  • I have a thread for health. I have had a lot of gut issues over the past 10 years and asking it to explain some of my symptoms has been very helpful not to replace the doctor but to help me bring up possible explanations to my doctors and get evaluations done. I plugged in all my symptoms, my health history, uploaded blood reports, and updated the AI in the same thread when I got results back or responses from doctors.

  • Now there is one last topic I wanted to bring up which is automation. AI not only has the ability to respond to chat, but has the ability to solve problems on its own and autonomously act like an employee what people are calling agents. With AI v1, this was self-driving cars. With AI v2, there's lots of tools out there. A few may have heard of OpenClaw? Controlling anything on your computer from Whatsapp? Now again these are very early tools I wouldn't recommend using without technical expertise as AI mistakes mean data can get deleted or actions can happen that you don't want, but it's important to know about because AI automation is becoming safer. For example, one of my friends setup an automation for my dad's pharmacy to confirm prescription refills. The agent determines which patients are likely to be out of refills soon and calls the patients on its own. It asks the patients if a prescription number is ready to be refilled and then marks it in the system for staff to start filling, saving time.

  • That's my AI v2 summary. Any questions?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment