Why Top Performers Are Building Third Brains With AI
Author Note: I’m particularly proud of this article. It’s a monster post that has been evolving since August. It represents my largest and most unique contribution on how to strategically chart your authentic, long-term course in the rapidly changing world of AI. Not only that, there are more article-specific perks for paid subscribers than any of the other posts.
Imagine having a team of low-cost brilliant employees who never sleep:
Your research team processes market data.
Your strategy unit models futures and identifies opportunities.
Your creative department generates and tests product variations.
Your marketing team creates content across all platforms simultaneously.
This is what AI-first knowledge workers, entrepreneurs, and thought leaders are building right now, but with AI employees rather than human ones.
For example, the founder and CEO of the largest company in the world is planning for 2,000 AI minds for every employee at his company. Tobi Lütke, founder of Shopify, recently used dozens of AI assistants working together over 200+ iteration cycles to create an entire keynote presentation from scratch.
As these AI-first pioneers build their AI workforce, they're discovering that deploying AI at scale requires three distinct systems…
Introducing The Three Brains Of AI
These interconnected “brains” each serve a unique cognitive function:
Your AI-first strategy (first brain): Deciding what to build with AI and how to build it.
Your data foundation (second brain): Organizing the information you and your AI workforce need. It acts as your "memory bank"—storing information, experiences, and knowledge.
Your mission control (third brain): Orchestrating AI agents to execute using that data. It functions as your "thinking patterns bank" (AI prompts)—storing ways of processing, analyzing, and creating.
This three-brain framework mirrors how our biological brains evolved and function:
Our prefrontal cortex handles executive functions like strategy and planning (first brain),
Our hippocampus manages memory storage and retrieval (second brain),
Our neural networks coordinate complex behaviors and responses (third brain).
Just as human civilization accelerated when we learned to augment our biological brain with external tools like writing and computers, we're now entering an era where systematically augmenting our thinking with AI will create similar exponential leaps in human capability.
With this understanding, let me show you how these three brains work together through a simple example—writing an article:
First Brain (Strategy): You decide to write an article on AI, targeting business leaders who want to stay ahead of the curve.
Second Brain (Data): You pull from your organized collection of research, personal experiences, expert interviews, and market trends you've saved over time. This isn't just bookmarks; it's a searchable knowledge base that grows smarter over time.
Third Brain (Mission Control): You create AI prompts for each step of the article creation process and connect each prompt to the relevant data from your second brain. For example, the steps might be:
Research
Article hook
Big idea for the article
Outline
Title
Intro
Writing
Editing
You Run The System. Once you complete the prompts, you manually go through each step in whatever your favorite model(s) are. Then, once you complete each step, you copy and paste the final output into the next prompt. If you have the wherewithal and technical expertise, the next step is to automate more and more parts of the process that are amenable to it.
The steps above represent a simple workflow. Given that I specialize in blockbuster articles, I’ve developed a much more in-depth process, which I write about in the article below:
Working together, these three brains transform what used to be a day-long writing process into something that takes just a couple of hours, with better results.
Finally, these three brains are interrelated on multiple levels:
Many prompts (third brain) must operate on stored information (second brain).
The transcripts of your conversations with AI can be stored back in the second brain as a form of “synthetic data”.
Both systems need similar organizational principles (tagging, search, versioning).
What You Get In Today’s Article
In this article, I provide a roadmap for creating, managing, and interconnecting these systems in detail for the first time ever on the Internet that I’m aware of. By the end of the article, you’ll have a road map to 5x your productivity in 2025 and 100x it in the next 2-3 years.
Imagine building a company where you get to focus on what you love as a solopreneur/small team while simultaneously getting the financial income and social impact of having a large company. That’s the promise of becoming an AI orchestrator.
Not only that, AI orchestrators are positioned for where the long-term of AI is moving. Rather than reactively being perpetually caught off guard, as an orchestrator, you can proactively skate where the puck is going.
Free Subscribers
Full 4,000+ word article
Paid Subscribers
Paid subscribers will get additional benefits focused on helping you quickly apply the ideas in the article and get real results:
Get access to my AI-First course in January: During the course, I go into further depth on the second brain and third brain and answer people’s questions about it live.
30-day prompt challenge. I’m tentatively planning a 30-day third brain challenge in March where participants create a new prompt every day of the month. I will cement plans for this in the coming weeks.
20+ Premium Prompts (coming in March): A collection of 20 pre-built premium prompts created by me to jumpstart your third brain.
My mindmap of the third brain I’m building: One of the hardest parts of getting started is getting ideas for what use cases to create prompts for. My mindmap will give you ideas and inspire you to create your own.
Deeper dive on the three brains: If you love the idea of an AI-first second brain and third brain, and you want to go a level deeper so you can take action, this section will help you.
11 provocative questions to ask your second brain to extract new insights. We now live in a world where we can ask any question to dozens of books, studies, articles, speech transcripts, interview transcripts, and articles. The hard part is now deciding what questions to ask. Below are vetted questions I love that will help you get new insights.
Second and third brain roadmap (more in-depth). Get a step-by-step action plan for transitioning through the stages from AI Tourist to AI Toolmaker to AI Orchestrator. I’m a big believer that a lot of success isn’t just about doing the right things, it’s about doing the right things in the right order. Doing the right things in the wrong order often leads to failure.
Premium prompt to turn knowledge into prompts. This prompt is explained later in the article. It is designed to turn any well-explained content you find online into a prompt.
Multi-disciplinary principles for a better third brain. Third brains pull from three fields—software design, business process automation, and second brains. Given that these fields have been around for a long time, there are a lot of lessons we can pull from them.
Overview of five second-brain and third-brain tools. I’ve spent dozens of hours researching, testing, selecting, and using tools to help me with my second and third brain. In this section, I share my top tools for AI second brain, prompt management, and multi-agent development.
In addition to these article-specific bonuses, you also get access to $2,500+ in other paid subscriber perks (courses, templates, prompts).
You get all of this for just $15/month making it one of the best deals in town…
Full Article
Let’s start by doing a deep dive into what a second brain and a third brain look like in an AI era.
As you read, keep in mind that what I present here is what I now see after 2,000+ hours focused on AI over the last two years. The intention isn’t to say that it’s where you should be now, but rather to inspire you with a vision and roadmap of what’s possible with AI so you start the journey today with confidence that all of your efforts will pay off.
How To Upgrade Your Second Brain In The AI Era
The Traditional Second Brain
Traditionally, a second brain is your digital, augmented memory—storing information, experiences, and insights.
Stated differently, it’s a knowledge management system that helps you manage information overwhelm by making it easy to store, organize, retrieve, use, and build upon the knowledge you’ve collected. With digital tools like Evernote, Notion, Obsidian, and Roam, it’s easier and more beneficial to create and use a second brain than ever. If you don’t know what a second brain is and want to go deeper, I recommend Tiago Forte’s Building A Second Brain book (free introductory guide).
These tools unlocked incredible potential, allowing us to tame information chaos. However, they remain limited to organizing and presenting what we feed into them, unable to bridge the gap to deeper creativity, interpretation, or entirely new ideas on their own.
What’s critical about this moment is that having AI inside your second brain isn’t just augmenting what a second brain is, it’s completely reinventing it. Specific AI features are coming online right now with profound, underexplored implications:
Collect 1,000x more data
Interact with your AI as a thought partner
Go beyond the data to generate novel insights
Turn data into prompts
AI Game-Changer #1: Collect 1,000x more data
AI is quickly trending toward infinite memory and free transcription, which makes it easier to collect data and then make sense of it with AI.
New software and hardware make it possible to capture transcripts from everywhere in your life and import them into your second brain:
Meetings
Phone conversations
Physical life audio (eg, PLAUD)
Video via your webcam
Audio notes you leave for yourself and others
Chat, calendar, documents, and emails
YouTube videos
Everything that happens on your computer screen (computer use)
Not only that, new tools like Otto, Exa, Gemini Deep Research, and Clay are making it easy to scrape the web for data and put it into your second brain. This means you’ll be able to autonomously collect info on:
Tools and use cases
Technological capabilities data
Industry data
Economic data
Cultural data (adoption, usage patterns)
Case studies
Trends
Inflection points
Paradigm shifts
Hacks
Protocols
Mental models
People
The implications of having so much data and the intelligence to process it are profound and underexplored:
Before, the second brain was about collecting snippets of information that we want to remember and use. With AI, we collect 1,000x more information that can also serve as context for AI conversations and AI prompts.
More of the knowledge collected in our second brain is non-personal information that would be valuable to other people’s second brains as well. Thus data in our second brain might actually become an asset we can sell.
When making decisions or solving problems, people will first ask 'What data would lead to the best decision?' and 'How can we collect that data?' rather than relying on immediately available information.
There are already harbingers of this future:
In my own business, I’m building a personal tool for web scraping that:
Checks for daily updates on 200+ blogs, podcasts, newsletters, X accounts, YouTube videos, and academic profiles.
Converts all the new media that was created into transcripts in a database.
Explore insights from that data.
Furthermore, new tools help you get amazing capabilities out of the box. For example, Google’s Deep Research tool does unlimited in-depth searches of 200+ websites and then creates in-depth articles with spreadsheets.
I provide a more in-depth breakdown of Deep Research in the paid section of this post. It includes two curated videos along with a step-by-step overview of how it works.
AI Game-Changer #2: Interact with your AI as a thought partner
In the past, you interacted with your second brain by retrieving relevant data, reading it, and then reflecting on it by yourself. Now, we have a whole new paradigm where we literally talk to an intelligence to deepen our understanding faster.
Not only that, you’ll be able to ask questions to your AI that no other being could answer, because you’ve shared more information with AI than anyone else. This enables you to extract new insights about yourself that go way beyond the capabilities of second brains of the past.
For example, you could explore contradictions in your thinking by asking:
“What contradictions exist in my thinking across these notes from the past year?”
The implications here are profound:
Instead of trying to remember facts, you might focus on remembering good questions to ask.
The relationship with your notes becomes more like a partnership than a tool.
When AIs can understand our thought patterns deeply, it creates a level of intellectual intimacy we've never experienced before.
All of this might lead to new forms of self-awareness and personal growth we can't yet imagine.
There are already harbingers of this future:
With NotebookLM (premium version), which I use regularly, you can upload up to 300 documents (books, studies, YouTube videos, and podcast transcripts) that are 375,000 words each. Then you can ask questions that extract unique insights, that also include exact citations. One of my favorite questions to ask to get things going is:
”When you synthesize the insights from all of these sources, what are the most profound and important emergent insights that arise that are not in any of the sources and have never been shared before by any human?”
I’ve spent a lot of time collecting the most valuable questions that one could ask a second brain. I share many of the most valuable ones in the paid section of the post.
AI Game-Changer #3: Go beyond the data to generate novel insights
AI isn’t just good at understanding our data, it’s good at using the data to form new insights that aren’t in the data or even talked about in any other data in human history. Stated differently, AI can perform a series of “cognitive operations” that build upon that data and create new insights. For example…
Expand upon
Highlight key parts
Summarize
Contextualize
Categorize
Abstract
Connect
Sequence
Prioritize
Extrapolate
Synthesize
Integrate
Go meta
While it might take a human many hours to perform any one of these operations, AI can do it in seconds. This is all a big deal because many people think that AI can’t form its own unique insights. I think people make this mistake because, for most people, the multi-step process for how humans form insights is a complete black box. But, when you understand the specific operations that a human brain does to form insights, you find that AI is good at performing many more of these than we think.
This has profound implications I haven’t seen anyone explore:
The data you create in conversation with AI becomes part of your second brain
This new data allows you to have even deeper insights for future conversations with AI
Which creates a virtuous loop of learning
For example, here’s what my process for researching a forthcoming article on Human Superintelligence looks like:
Collect relevant resources across fields (books, video transcripts, studies, etc)
Upload them into my second brain (eg, NotebookLM in this case)
Extract key info to create a research brief so I can see the big picture (key concepts, thinkers, excerpts, timeline of ideas, mental models, opportunities to increase human intelligence, open questions)
Do deep analysis to generate novel ideas (synthesize the ideas into unique insights not written about in any of the resources)
Shown visually…
Data Foundation. This is the data you input and import into your second brain.
Research Brief. This is a document you can create to summarize all of the sources on a topic in order to get a consolidated overview of it. The research brief can also be a data source for a prompt.
Analysis. The analysis step has AI analyze the research brief from multiple perspectives and generate emergent insights that aren’t in the data you uploaded.
Ideation. From these insights, you can coin new Trademark Ideas that you become known for.
I already see the harbinger of this process in my own life:
I used to painstakingly read tons of books where most of my time was spent simply finding the useful information in them and then transferring them to my second brain.
It used to take incredible amounts of time to manually generate unique insights from everything I was absorbing.
Now, once I have the data in my system, I spend a lot of days exploring mind-blowing insights with AI that I’ve never seen written about anywhere else. It’s surreal.
I’m just as actively involved in everything, but I’m involved at a much higher level that is more impactful.
And, I know that I’m still just scratching the surface. Large parts of what I do will be automated soon.
AI Game-Changer #4: Automatically turn data into prompts
When I take a class or watch a YouTube video explaining something really well with examples, I don’t just save it to my second brain; I use that as a data source to create a prompt.
To make this more concrete for you, I found a really good 5-minute video clip from a Nicolas Cole video on how to create tangible offerings for a paid newsletter post:
Source: Nicolas Cole
After finding the video, I used the Glasp Chrome plugin to get the YouTube transcript. Next, I turned the transcript into a Claude prompt such that whenever I’m writing an article, I copy the article into the prompt, and Claude gives me suggestions on creating a paid article companion; finally, it then actually creates the first version of it for me.
I share this prompt for paid subscribers at the bottom of this post.
This is more profound than it appears:
When people teach things online or write articles, they provide the exact data needed to create effective prompts. This includes metaphors, examples, clear instructions, nuances, and more.
Taking things to the next level, with a prompt that identifies “prompt creation” opportunities, it becomes possible to:
Copy and paste a resource (article, class) into AI
A first prompt identifies multiple “prompt creation” opportunities in the resource
A second prompt turns each opportunity into a separate prompt
Thus, it becomes possible to easily and automatically create hundreds of prompts that form the data foundation for your third brain.
This changes the definition of what it means to be a student. Rather than just doing exercises, you may also create and use prompts based on the resource.
Bottom line:
The concept of a "second brain" is evolving. What started as a personal knowledge management system—a place to store notes and ideas so you can retrieve and use them later—is transforming into something more powerful: the data infrastructure that also powers your AI workforce:
Yesterday: Personal library of notes, articles, and insights.
Today: Organized information that enables you and AI agents to work effectively.
Tomorrow: Dynamic knowledge ecosystem that grows smarter with interaction.
Next, what’s key to understand is that the value of the data isn’t just the data by itself. It’s how the data forms the foundation for your third brain…
This Is What A Third Brain Looks Like In The AI Era
Your third brain is your command center—managing hundreds (and eventually thousands) of specialized AI agents, each trained through carefully crafted prompts to handle specific aspects of your workflow.
It’s valuable to save your prompts for several reasons:
It prevents you from creating the same prompt from scratch over and over.
It creates a foundation to refine the prompt so its responses are more valuable and reliable.
Eventually, with an advanced third brain, it allows you to sell your prompts or automate them so they can work without you.
And like a second brain, things like search, folders, and tagging make it easy to find the right prompts when you need them.
Said differently, you can think of the third brain as a prompt management system that helps you manage thinking overwhelm (prompt overflow) by making it easy to build a library of AI prompts and use them individually or together in the right context.
The most successful companies of the past built sophisticated systems to manage human employees. The leaders of tomorrow are building third brains to manage their AI workforce. The difference between using AI occasionally and deploying it at scale is the difference between having a few freelancers versus running a full organization. One requires simple tools; the other demands management systems.
What A Basic Third Brain Includes
At a fundamental level, a basic third brain is a library of:
Very simple few-sentence, good enough prompts
Where each prompt represents a helpful use case that is valuable to you
In my experience, the first time you find a helpful prompt that you think could be helpful in the future, you should save it to your prompt library. Said differently, it’s better to be safe and create a prompt you might use in the future than to hold off.
The reason I recommend this is that when I save a new prompt, I find that it activates my reticular activation system. Therefore, I’m more likely to remember and use it when I need it. Not only that, as soon as I start using a prompt regularly, I find that I start improving it.
As a starting point, you can save the prompt in the tool you use the most:
What An Intermediate Third Brain Includes
The intermediate third brain is a larger library of more developed prompts you repeatedly use and find helpful.
With more advanced prompts, you might include sections like:
Goal of the prompt
Format and style of the prompt’s output
Data source for the prompt
Before and after examples of what the prompt should do
Audience
Etc.
What An Advanced Third Brain Includes
As you create more and more prompts, you start to get more distinctions that allow you to operate with higher levels of productivity. For example, you follow some of the protocols below:
Create different types of prompts. You realize that the prompts fall into different categories. Some help you identify opportunities to create prompts. Other execute tasks. Others test and improve outputs of other prompts. And so on.
Store prompt snippets. You also realize that there are smaller, reusable components within prompts that you find helpful in more than one prompt. By separating these snippets from individual prompts, it’s possible to easily add them to new prompts and update them once centrally in order for it to update across all prompts. Thus, you now also have a library of prompt building blocks that accelerate future development.
Use an Independent Prompt Environment. Most people either store all of their prompts on one platform, or store their AI prompts haphazardly across different platforms. They might have some saved in ChatGPT, others in Claude, and still more scattered across various notes and documents. This approach breaks down once you start having multiple versions of prompts on different platforms. Eventually, it makes more sense to just use one independent prompt environment that integrates with all of the AI models.
Use version control. As prompts get more complex, small changes could have unexpected consequences you don’t realize until the future. Therefore, it’s helpful to keep old versions of prompts stored in such a way that you can easily access them.
In the paid section at the bottom of this post, I go much deeper on prompt types and prompt snippets along with independent prompt environments to consider. This will save you hours of work and pain dealing with avoidable friction.
What A Master Third Brain Includes
The final step is automating as much as you can or want so that the prompts can act autonomously to get things done. At this stage, your prompt library turns into a mission control.
First, it manages your AI agents. Each prompt in your system is like an employee with specific skills and responsibilities. Some are specialists, trained for particular tasks like market analysis or content creation. Others are generalists, able to handle a variety of basic tasks. Your mission control system helps you deploy the right agent for each job.
Second, it orchestrates workflows. Complex projects often require multiple AI agents working together, just like complex business projects require multiple employees collaborating. Your mission control system manages these workflows, ensuring smooth handoffs between agents and maintaining quality throughout the process.
This means being able to chain prompts together, establish clear handoff points between agents, and build quality checks at critical stages. Below are some great visuals of what more advanced workflows can look like courtesy of Agent Recipes.
Important note: Don’t worry about the specifics of what each box means at this point. Just think of each box as a separate prompt in your prompt library. 1) Prompt chaining:
2) Orchestration
3) Evaluator-optimizers
4) Autonomous agent
Third, it drives continuous improvement. Every time an AI agent completes a task successfully, that process can be refined and standardized. Every time there's a failure, lessons can be learned and incorporated. Your mission control system captures these learnings and helps you build increasingly sophisticated workflows.
Fourth, it integrates with your second brain. Your AI workforce needs to connect with your data foundation, access necessary tools and resources, and plug into your existing work systems. Your mission control system manages these connections, ensuring smooth operation across your entire AI-first operation.
As this AI-first future unfolds, a new divide is emerging among knowledge workers where top, AI-enhanced performers will gain the ability to be 1,000x more productive than good performers…
The New AI Divide
This divide is not just between those who use AI and those who don't—it's between three distinct groups with dramatically different approaches to AI deployment:
AI Tourists
AI Toolmakers
AI Orchestrators
These aren't just different types of AI users—they're stages in a natural progression that all knowledge workers will go through. Almost everyone starts as an AI Tourist, experimenting with different tools and discovering what's possible. Some advance to become AI Toolmakers, building their first reusable systems, and a growing number are evolving into AI Orchestrators, developing comprehensive systems to manage their AI workforce.
Each level requires a significant increase in time investment and delivers an exponential increase in value.
Those who are most at-risk are those who stop improving or don’t improve at a fast enough rate compared to the market.
1. AI Tourists
Overview: AI Tourists use AI reactively—writing emails, summarizing documents, and solving problems as they arise.
Return On Investment:
Some protection from being disrupted by AI
Increased productivity
Time Investment: 1-5 hours per week.
Bottleneck: Start from scratch with each new task.
Mental Blocks:
Overconfidence: They often think they are being proactive because they’re using and benefiting from AI regularly while many people they know aren’t, but they’re just tapping into the surface level of productivity.
Overwhelm: These individuals listen to a handful of thought leaders, don’t have the time to understand what’s happening at a fundamental level, and progress slowly to avoid overwhelm.
2. AI Toolmakers
Prompt collection: AI Toolmakers have moved beyond one-off prompts. They build small collections of proven AI solutions—maybe a dozen carefully crafted prompts they reuse for common tasks.
Return on Investment: They're seeing greater returns from AI, but their approach still treats each prompt as a standalone tool rather than part of an integrated system.
Bottleneck: They need to manually copy and paste input and outputs between the right prompts.
Time Investment: 10-15 hours per week
3. AI Orchestrators
These are the pioneers who are all in on building AI-first operations. They don't just use AI tools; they create and manage entire AI workflows as AI employees. Their prompts aren't isolated instruments but components of a larger system. They're building true command centers for their AI workforce, with clear processes for creating, testing, deploying, and refining their AI capabilities.
We get a glimpse of what AI orchestration might look like in the following video clip of Shopify founder and CEO Tobi Lütke explaining how he used dozens of agents and hundreds of communication loops between them to create his keynote address and slides for the company’s annual employee conference:
I write about Lutke’s system in-depth in AI-First Billionaire CEO: Building Your Own AI Dream Team Is Your Greatest Career Move.
Return On Orchestration (ROO)
Automation: AI orchestrators provide a major leap forward above AI toolmakers, because of automation. Thus, they remove themselves as the bottleneck.
Scale: AI orchestrators can manage AI workforces that are larger and larger gradually increasing from a few AI employees to dozens to hundreds to thousands.
Productivity: One AI orchestrator can have 1,000x productivity compared to an AI tourist.
Conclusion
The performance gap between these three groups is widening rapidly. While AI Tourists are still figuring out basic prompts and AI Toolmakers are building their collections, AI Orchestrators are developing comprehensive systems that multiply their capabilities 100x and soon 1,000x.
The good news?
The key is recognizing where you are in this journey and taking intentional steps toward greater mastery. Whether you're just starting to explore AI or already building sophisticated systems, understanding these stages helps you plot your path forward. What I personally love about this roadmap is that it gives me a way to chart my path forward in an uncertain AI world in a way that is both authentic and sustainable rather than reactive and threat-based.
When you take the time to develop these three brains – your vision, your knowledge foundation, and your execution system – you create space for the work that truly matters. The mechanical tasks fade into the background while your creativity and insight take center stage. And that's the real opportunity here: not just to be more productive, but to focus your energy on the problems worth solving, the ideas worth exploring, and the changes worth making.
The tools are here. The path is clear. The only question is: are you ready to build something remarkable?
Now that you understand the three brains at a basic level, below are perks just for paid subscribers to help you take action…
10 PAID PERKS:
APPLY THE IDEAS FROM THE ARTICLE
AI-First course in January: During the course, I go into further depth on the second brain and third brain and answer people’s questions about it live.
30-day prompt challenge (coming in March): I’m tentatively planning a 30-day third brain challenge in March where participants create a new prompt every day of the month. I will cement plans for this in the coming weeks.
20+ Premium Prompts (coming in March): A collection of 20 pre-built premium prompts created by me to jumpstart your third brain.
My mindmap of the third brain I’m building: One of the hardest parts of getting started is getting ideas for what use cases to create prompts for. My mindmap will give you ideas and inspire you to create your own.
Deeper dive on the three brains: If you loved the idea of an AI-first second brain and third brain, and you want to go a level deeper so you can take action, this section will help you.
11 provocative questions to ask your second brain to extract new insights. We now live in a world where we can ask any question to dozens of books, studies, articles, speech transcripts, interview transcripts, and articles. The hard part is now deciding what questions to ask. Below are vetted questions I love that will help you get new insights.
Second and third brain roadmap (more in-depth). Get a step-by-step action plan for transitioning through the stages from AI Tourist to AI Toolmaker to AI Orchestrator. I’m a big believer that a lot of success isn’t just about doing the right things, it’s about doing the right things in the right order. Doing the right things in the wrong order often leads to failure.
Prompt to turn knowledge into prompts. Earlier in the article, I mentioned how I turn really good explanations into prompts. My full prompt appears below.
Multi-disciplinary principles for a better third brain. Third brains pull from three fields—software design, business process automation, and second brains. Given that these fields have been around for a long time, there are a lot of lessons we can pull from them.
Overview of five second-brain and third-brain tools. I’ve spent dozens of hours researching, testing, selecting, and using tools to help me with my second and third brain. In this section, I share my top tools for AI second brain, prompt management, and multi-agent development.