Agree 100%. For me, it’s about spaces for the team. So not just mg nerdy brand knowledge but a product marketer’s insights alongside a growth marketer’s channel expertise. I built this in Claude Projects, as team had corporate seats. It’s a collective brain that can execute fast, without compromising quality.
Yes, it’s just wild how everything will easily connect when we put certain knowledge altogether. It makes every information flows smoothly and AI gets 10x smarter with provided knowledge we feed them to.
The mindset helps me to record everything and feed them to AI everytime I learn something new.
This is what I needed to hear. I’m overwhelmed with all the tools and love each of them for different reasons. This gives me a plan moving forward. Thank you! 🙏
Great post. I'm still grappling with these issues -- AI tools offer so many possibilities, but crafting a personal system for oneself is still a project.
It's becoming clear that with all the brain and consciousness theories out there, the proof will be in the pudding. By this I mean, can any particular theory be used to create a human adult level conscious machine. My bet is on the late Gerald Edelman's Extended Theory of Neuronal Group Selection. The lead group in robotics based on this theory is the Neurorobotics Lab at UC at Irvine. Dr. Edelman distinguished between primary consciousness, which came first in evolution, and that humans share with other conscious animals, and higher order consciousness, which came to only humans with the acquisition of language. A machine with only primary consciousness will probably have to come first.
What I find special about the TNGS is the Darwin series of automata created at the Neurosciences Institute by Dr. Edelman and his colleagues in the 1990's and 2000's. These machines perform in the real world, not in a restricted simulated world, and display convincing physical behavior indicative of higher psychological functions necessary for consciousness, such as perceptual categorization, memory, and learning. They are based on realistic models of the parts of the biological brain that the theory claims subserve these functions. The extended TNGS allows for the emergence of consciousness based only on further evolutionary development of the brain areas responsible for these functions, in a parsimonious way. No other research I've encountered is anywhere near as convincing.
I post because on almost every video and article about the brain and consciousness that I encounter, the attitude seems to be that we still know next to nothing about how the brain and consciousness work; that there's lots of data but no unifying theory. I believe the extended TNGS is that theory. My motivation is to keep that theory in front of the public. And obviously, I consider it the route to a truly conscious machine, primary and higher-order.
My advice to people who want to create a conscious machine is to seriously ground themselves in the extended TNGS and the Darwin automata first, and proceed from there, by applying to Jeff Krichmar's lab at UC Irvine, possibly. Dr. Edelman's roadmap to a conscious machine is at https://arxiv.org/abs/2105.10461, and here is a video of Jeff Krichmar talking about some of the Darwin automata, https://www.youtube.com/watch?v=J7Uh9phc1Ow
Nice post. In our AI training sessions for B2B marketers, we often see this challenge—AI tools offer enormous potential, but building a system that fits your workflow takes real strategy. We recommend starting with a minimum viable toolset that supports your core day-to-day needs.
yes, i think all goes back to our daily workflow. you dont need more tools, you just need fewer tools but they complement and speak to each other, thats where major impact comes from.
It’s the starting over each time that really gets me.
I’ve been using GPT Projects more and more for that reason, though not nearly as extensively as what you’ve shared here. Using Notebook LM to extract patterns and key takeaways, then feeding that back into the project, is absolute gold.
It takes time and effort to do it well, but the payoff is worth it. Thanks for sharing this I really appreciate the insight.
My question is whether we need to switch back and forth between so many tools? I have been using mostly Claude (often with the lex.page interface for writing). But when OpenAI added image generation and memory, I decided to experiment with it. This week I had two use cases where only o3 was able to deal with my questions.
Now there's all the Claude 4 announcements, and I feel like I need to go check that out. The Red Queen effect is killing me.
This isn't just a budgeting question--it's a feeling that I need a home. You seem very comfortable bopping back and forth between all the models. I'd love to know more about why--or is it "just" curiosity and FOMO? Perfectly understandable under the circumstances.
I can totally relate with this and feel the pain of switching tools :)
Thanks for sharing.
I think it goes back again to tools that suit with your workflow.
I mainly use Claude for its project knowledge for writing, coding, and MCPs.
But I still go to ChatGPT for generating image and doing deep research with its o3 model mainly.
And Gemini 2.5 pro for planning and reasoning when I'm brainstorming ideas and building development plan for when I'm vibe coding my apps, depending which projects I'm working on.
I also have FOMO with Claude 4 but so far Claude is best at coding. So nothing has changed for me, it's just get better.
Every model has their own unique strength. The key is to know which one that does the job well for you and stick to it. But, I guess we are still in the early growth of AI, so having to try multiple models and figure out which one that suits our needs still required.
Yes ChatGPT knows but for me it's still small fragment of it and it's still missing a lot of context and expertise.
Project knowledge both GPT and Claude can offer something deeper about specific challenges or domains in your life/work and can be used across multiple usage depending what information you feed them to. It's like building the augmented AI brain.
For example: you can clone yourself as a manager and use CustomGPT to give feedback to your teammates. I dont think u can do that with memory.
Hell yes, Wyndo! Thank you for demonstrating not just what these tools can do, but how they can be integrated in a way that's greater than the sum of their parts. You're showing what it can actually mean to be "AI-first" rather than just throwing it around as a buzz word.
The next big thing I guess how to make all these interconnection run smooth and sync automatically so we can become truly AI-first brain augmented in our day to day lives!
MCP and AI agent are going to be game-changer for this!
This is a good mindset. I’ve recently started building it to support my writing with Claude projects and Notebook LLM. It began with the simple need to avoid repeating my desired style requirements. During the writing process, I actively note or ask AI about the things I tend to repeat, then include them in the style guide. This saves time (less editing) and significantly improves quality.
By the way, I hope you’ll share more articles about vibe coding, like how to go from a PRD and guide Cursor to write code based on it.
Great to know that you’ve been using NLM and Claude project: game-changer!
Vibe coding has been sitting on my backlog for so long. But somehow I’ve been postponing this. Will get it in the next 2-3 weeks. Wanted to find unique angles as non-coders guide that people can follow. Thanks for surfacing this Luan!
Totally resonate with your frustration around starting over with a new project!
I really loved your suggestion to document things in any format and keep feeding it back into AI.
Personally (and lazily 😅), I just use ChatGPT and its project feature for storing history. Even so, having that preservation is incredibly helpful, it keeps AI aligned with us, not just some generic theme.
Grateful that ChatGPT remembers conversations now, I’m hoping it’ll soon be able to cross-check chat history too. Feels like we’re getting close.
This article really hits the mindset shift that AI users need. Thanks for sharing this great work!
Interestingly, Notion is looking to bring all this inside their platform also. Their latest AI releases have been impressive, but I’m already (habit-wise) used to opening up GPT or Claude. Something to be said about having separate UIs for separate tasks too 👀
I've a question: going by this post, do you think it's necessary to invest on the tools you've mentioned? Or, is it ok to make do with the tools we currently have acccess (free or otherwise)? I know you've said it's not about the tools, but thought I'd ask for an elaborate clarification if possible.
what tools currently u are using? I think u can just subscribe $20/month on either claude/chatGPT and those have answered the problem. NotebookLM is free so no worry about it. If you want to explore Gemini, it's also free.
The workflow FOMO is sooo real.
Your idea of AI environments is a great reframe. Love how you’re not just using tools, but building systems that evolve with your thinking.
Thanks for sharing the behind-the-scenes. It’s great to see how intentionally you’ve shaped your process.
It's good for our sanity rather than always chasing new tools.
Being more intentional makes me realize which tools that actually matter so I dont FOMO myself into the abyss 😅
Agree 100%. For me, it’s about spaces for the team. So not just mg nerdy brand knowledge but a product marketer’s insights alongside a growth marketer’s channel expertise. I built this in Claude Projects, as team had corporate seats. It’s a collective brain that can execute fast, without compromising quality.
Yes, it’s just wild how everything will easily connect when we put certain knowledge altogether. It makes every information flows smoothly and AI gets 10x smarter with provided knowledge we feed them to.
The mindset helps me to record everything and feed them to AI everytime I learn something new.
Cool I’ve subscribed!
https://aimaker.substack.com/p/i-stopped-chasing-ai-tools-and-started-building-ai-spaces-here-what-i-learned/comment/119296561?r=b8pvb&utm_medium=ios
This is what I needed to hear. I’m overwhelmed with all the tools and love each of them for different reasons. This gives me a plan moving forward. Thank you! 🙏
Glad u find it useful. Everything starts from understanding workflow then it evolves into space and system that surrounds us with more capabilities
Great post. I'm still grappling with these issues -- AI tools offer so many possibilities, but crafting a personal system for oneself is still a project.
Yes it takes time. Take it slowly, but the return is so damn amazing when u can use it as system rather than workflow.
It's becoming clear that with all the brain and consciousness theories out there, the proof will be in the pudding. By this I mean, can any particular theory be used to create a human adult level conscious machine. My bet is on the late Gerald Edelman's Extended Theory of Neuronal Group Selection. The lead group in robotics based on this theory is the Neurorobotics Lab at UC at Irvine. Dr. Edelman distinguished between primary consciousness, which came first in evolution, and that humans share with other conscious animals, and higher order consciousness, which came to only humans with the acquisition of language. A machine with only primary consciousness will probably have to come first.
What I find special about the TNGS is the Darwin series of automata created at the Neurosciences Institute by Dr. Edelman and his colleagues in the 1990's and 2000's. These machines perform in the real world, not in a restricted simulated world, and display convincing physical behavior indicative of higher psychological functions necessary for consciousness, such as perceptual categorization, memory, and learning. They are based on realistic models of the parts of the biological brain that the theory claims subserve these functions. The extended TNGS allows for the emergence of consciousness based only on further evolutionary development of the brain areas responsible for these functions, in a parsimonious way. No other research I've encountered is anywhere near as convincing.
I post because on almost every video and article about the brain and consciousness that I encounter, the attitude seems to be that we still know next to nothing about how the brain and consciousness work; that there's lots of data but no unifying theory. I believe the extended TNGS is that theory. My motivation is to keep that theory in front of the public. And obviously, I consider it the route to a truly conscious machine, primary and higher-order.
My advice to people who want to create a conscious machine is to seriously ground themselves in the extended TNGS and the Darwin automata first, and proceed from there, by applying to Jeff Krichmar's lab at UC Irvine, possibly. Dr. Edelman's roadmap to a conscious machine is at https://arxiv.org/abs/2105.10461, and here is a video of Jeff Krichmar talking about some of the Darwin automata, https://www.youtube.com/watch?v=J7Uh9phc1Ow
I really enjoyed this! Love the idea of a cohesive environment than a linear workflow. I appreciate this will refer to this again!
Glad u find it useful :)
Nice post. In our AI training sessions for B2B marketers, we often see this challenge—AI tools offer enormous potential, but building a system that fits your workflow takes real strategy. We recommend starting with a minimum viable toolset that supports your core day-to-day needs.
yes, i think all goes back to our daily workflow. you dont need more tools, you just need fewer tools but they complement and speak to each other, thats where major impact comes from.
Your comment is noted so I can explore your Substack.
Wyndo posts really great information. Do you agree?
It’s the starting over each time that really gets me.
I’ve been using GPT Projects more and more for that reason, though not nearly as extensively as what you’ve shared here. Using Notebook LM to extract patterns and key takeaways, then feeding that back into the project, is absolute gold.
It takes time and effort to do it well, but the payoff is worth it. Thanks for sharing this I really appreciate the insight.
I'll be waiting NLM to open their MCP/API so we can connect it through n8n/gumloop and let AI agent does the rest while we chill :)
"I've surrounded myself with AI tools that complement each other rather than competing for my attention. "
So much this! I'll be subscribing. I feel heard!
Very useful approach--thanks!
My question is whether we need to switch back and forth between so many tools? I have been using mostly Claude (often with the lex.page interface for writing). But when OpenAI added image generation and memory, I decided to experiment with it. This week I had two use cases where only o3 was able to deal with my questions.
Now there's all the Claude 4 announcements, and I feel like I need to go check that out. The Red Queen effect is killing me.
This isn't just a budgeting question--it's a feeling that I need a home. You seem very comfortable bopping back and forth between all the models. I'd love to know more about why--or is it "just" curiosity and FOMO? Perfectly understandable under the circumstances.
I can totally relate with this and feel the pain of switching tools :)
Thanks for sharing.
I think it goes back again to tools that suit with your workflow.
I mainly use Claude for its project knowledge for writing, coding, and MCPs.
But I still go to ChatGPT for generating image and doing deep research with its o3 model mainly.
And Gemini 2.5 pro for planning and reasoning when I'm brainstorming ideas and building development plan for when I'm vibe coding my apps, depending which projects I'm working on.
I also have FOMO with Claude 4 but so far Claude is best at coding. So nothing has changed for me, it's just get better.
Every model has their own unique strength. The key is to know which one that does the job well for you and stick to it. But, I guess we are still in the early growth of AI, so having to try multiple models and figure out which one that suits our needs still required.
Do you use a subscription of ChatGPT?
I feel like a lot of the advantages you mention here of an environment I get already within my subscription of ChatGPT with its memory features.
ChatGPT frequently draws on our past conversations and my domains of mastery, even without me asking it to.
Wondering if Claude projects offers additional advantages
I have.
Yes ChatGPT knows but for me it's still small fragment of it and it's still missing a lot of context and expertise.
Project knowledge both GPT and Claude can offer something deeper about specific challenges or domains in your life/work and can be used across multiple usage depending what information you feed them to. It's like building the augmented AI brain.
For example: you can clone yourself as a manager and use CustomGPT to give feedback to your teammates. I dont think u can do that with memory.
Hell yes, Wyndo! Thank you for demonstrating not just what these tools can do, but how they can be integrated in a way that's greater than the sum of their parts. You're showing what it can actually mean to be "AI-first" rather than just throwing it around as a buzz word.
Glad u resonated well with this Takim :)
The next big thing I guess how to make all these interconnection run smooth and sync automatically so we can become truly AI-first brain augmented in our day to day lives!
MCP and AI agent are going to be game-changer for this!
This is a good mindset. I’ve recently started building it to support my writing with Claude projects and Notebook LLM. It began with the simple need to avoid repeating my desired style requirements. During the writing process, I actively note or ask AI about the things I tend to repeat, then include them in the style guide. This saves time (less editing) and significantly improves quality.
By the way, I hope you’ll share more articles about vibe coding, like how to go from a PRD and guide Cursor to write code based on it.
Great to know that you’ve been using NLM and Claude project: game-changer!
Vibe coding has been sitting on my backlog for so long. But somehow I’ve been postponing this. Will get it in the next 2-3 weeks. Wanted to find unique angles as non-coders guide that people can follow. Thanks for surfacing this Luan!
Totally resonate with your frustration around starting over with a new project!
I really loved your suggestion to document things in any format and keep feeding it back into AI.
Personally (and lazily 😅), I just use ChatGPT and its project feature for storing history. Even so, having that preservation is incredibly helpful, it keeps AI aligned with us, not just some generic theme.
Grateful that ChatGPT remembers conversations now, I’m hoping it’ll soon be able to cross-check chat history too. Feels like we’re getting close.
This article really hits the mindset shift that AI users need. Thanks for sharing this great work!
I think this is the general problem with AI: low context window and lack of interconnected apps capability.
If only connecting multiple apps can be easier, a lot of problems will be easily solved.
But this can be solved with AI automation tools and MCPs.
Though the current tech still not perfect, we'll arrive there soon enough!
Spot on.
Interestingly, Notion is looking to bring all this inside their platform also. Their latest AI releases have been impressive, but I’m already (habit-wise) used to opening up GPT or Claude. Something to be said about having separate UIs for separate tasks too 👀
Same!
I think what u can do with Notion is to use MCP and connect it with Claude so it can access ur notes and let Claude takes full control over ur Notion.
I've a question: going by this post, do you think it's necessary to invest on the tools you've mentioned? Or, is it ok to make do with the tools we currently have acccess (free or otherwise)? I know you've said it's not about the tools, but thought I'd ask for an elaborate clarification if possible.
what tools currently u are using? I think u can just subscribe $20/month on either claude/chatGPT and those have answered the problem. NotebookLM is free so no worry about it. If you want to explore Gemini, it's also free.