The task today was huge for me. It closed a loop that was the basis for me starting to learn programming and computer skills. I am going to share a LinkedIn post I made a while back that explains it best. This was before AI was what it is today. I started my career in the trades, and at that time there was not a strong focus on computer skills. I come from a generation where computers were only just starting to appear in schools. We learned the basics, but technology did not have the importance it does today. When I returned to school in 2008, my computer skills were barely functional, but enough to get me through my Materials Engineering Technology studies. When I joined my current employer, I will be honest: my office-related computer skills were pretty weak. Like many people, I had “Microsoft Office skills” on my resume, but in reality I was just getting by. At that point, continuous learning was not a major focus for me. A few years later, our department hired a new employee named Praveen. Praveen had incredible computer skills, and working with him opened my eyes to what was possible. One moment in particular has always stuck with me. At the time, I was working full time while also carrying a full-time course load for my BTech. One lunch hour, I was manually transferring warranty claim data from multiple vehicle manufacturers into an Excel sheet for a systems reliability project. By my estimate, it was going to take me somewhere between 16 and 20 hours just to build the data set. Praveen came over and asked what I was doing. I explained the project and how long I thought it would take. He told me he thought he could do it by the end of lunch. I honestly thought he was dreaming. He explained web scraping and how he would approach it. I will spare everyone the technical details, but by 12:50 he had sent me a fully formatted Excel file with the entire data set I needed for my project. He had saved me days of work. I told him I would give him $100 if he could actually pull it off. He tried to refuse it, but I made him take it.

I completed day 1 of the AIS challenge over the weekend. Had to engage with the community to get to level 2 to post it. Over the past year or so I have spent a lot of time working with AI but mostly in the chat interfaces so essentially a super powered chat bot. Focused a lot of prompt engineering and context which has been a real help. Built some really neat frameworks for teaching and for translation projects but those frameworks always had to be what was entered first. The last project I did in this method took me about 30 hours. After completing day 1 of this challenge I realize that same project could have been less than 2 hours and automated. It is incredible how quickly I have been learning these skills. Doing that previous project in 30 hours labeled me as one of the AI experts at my work, now I feel how poor my skills were at that time but were still well above my peers there. I had some issues with this project, mostly because automation and API calls etc were new to me. I used Claude chat a lot to help me through some of the stages and iterated on the final product a lot. I set everything up per the lesson but the infographics were sub standard so I switched up the source of those for the final product. I still need to work on branding for myself, but because I am starting a consulting business I need to do that as a side project and decided to keep this more towards learning the key concepts. The newsletter to someone that doesn't work in the industry probably sounds pretty good, but as an expert it lacks a bit of depth that would come with context and my anti AI language skills. Overall the lesson was very informative and I definitely see a ton of value in the skills I learned and am already applying them to other areas. As a side note, if you can't get bypass permissions to work you need to run VSCode as administrator. I was confused why it kept asking me for approvals and finally I tried running as administrator and everything worked properly.

Can anyone give me some recommendations on how to build a good AI-powered “second brain” that will help me conduct powerful research and save on tokens when consulting with AI? Or share your experiences creating a “second brain” and using it in your daily life?

I see so many people here in completely different stages…but everyone talks like they’re at the same level. Let’s make it real for a second. 🔥Which stage are you at right now? Stage 1 – Watching & learning 🫡🫡 You binge YouTube, Skool posts, and courses. You know a LOT in theory, but no real offers, no outreach, no clients yet. Stage 2 – Building & testingYou build workflows, AI agents, demos. Maybe tested on friends or fake leads. Still no proper paying client. Stage 3 – First clients inYou’ve closed 1–3 clients (even low-ticket). You’re delivering, fixing bugs, and realizing what actually matters. Stage 4 – Dialing it inYou have a clear niche + offer, 5+ clients, and now the problem is systems, fulfillment, and time. How to reply (copy this):Stage: 1 / 2 / 3 / 4Biggest struggle right now: Drop your answer in the comments 👇I’ll read through and reply with one thing I’d focus on if I were in your stage. 💎Me (so you know who’s asking): I’m Kartik, building BlitzFlow AI.I help service businesses (like HVAC )stop losing money from missed calls, slow lead response, and dead CRM leads using AI voice reception, speed-to-lead workflows, and automations.

I have to admit that I’m a bit confused by everything that has recently emerged around building an AI Operating System (AI OS). Let me explain. First, I came across the concept of a Wiki LLM, as described by Andrej Karpathy, and that seemed fairly straightforward to me. Then Graphiti entered the scene, introducing the idea of building knowledge graphs from complex relationships. However, it’s still not entirely clear to me whether Graphiti can be integrated into a Wiki LLM system to analyze complex resources such as Excel files, or whether it should be treated as a standalone system. And then there’s the AI OS concept itself. I’m not sure whether it’s intended to consolidate all these innovations into a single framework, or whether it represents an entirely different approach. I’m still at the beginning of my journey, and I’d like to build my own “second brain” in the most effective way possible, starting with the right tools from day one. Do you have any suggestions or recommendations on how to approach this?