🤝 Stop losing your best ideas
Dear curious mind,
Most of us treat our writing like a one-way street: we create, we publish, and we move on. But what if every newsletter, blog post, or article you've written could become a valuable source for future work? This week's shared insight explores how AI can help you transform your scattered writing archive into a searchable, reusable library of your own expertise.
Ready to unlock the hidden potential in your content archive? Let's dive in.
In this issue:
💡 Shared Insight
Your Writing Archive is an Untapped Goldmine
📰 AI Update
LM Studio: Now Free for Professional Use
Kyutai TTS: Open-Weight Voice AI That Outperforms ElevenLabs
🌟 Media Recommendation
Podcast: Simon Sinek on Finding Meaning When Life Feels Overwhelming
💡 Shared Insight
Your Writing Archive is an Untapped Goldmine
At March’s PKM Summit, I had a fascinating conversation with Bob Doto, author of the book "A System for Writing," about the Zettelkasten method. During our discussion, Bob shared an insight that seemed obvious once said but had never occurred to me: you can use your own writing as a source for future work.
This might sound like a small realization, but it's actually profound. Every article you write, every newsletter you publish, and every blog post you create becomes part of your personal library. You're not just creating content for your audience; you are building a repository of your own thoughts, insights, and expertise that can fuel future projects.
The challenge is that many publishing platforms, including Substack, where I publish this newsletter, aren't designed to make your content easily searchable or reusable for your own purposes. Your writing gets locked away in HTML format, scattered across individual posts that are difficult to work with systematically.
This is where AI becomes your ally. Inspired by Bob's insight, I decided to extract all my newsletter content and transform it into a more useful format. Substack provides an export feature that creates a zip file of all your posts, subscribers, and some analytics. Using Claude Code as a programming assistant, I converted my HTML-formatted Substack posts into clean Markdown files. Why Markdown? Because it's the perfect format for feeding into AI models: clean, structured, and easy to process. If you want to do the same, here is the link to my GitHub repository, which converts all pages in a Substack export to markdown files.
Now I have all my past writing organized in a way that lets me ask AI, "What have I already written about voice interfaces?" or "Find all my insights about privacy in AI." Instead of manually searching through dozens of posts, I can instantly surface relevant content I've already created.
But this goes beyond just finding old content. By feeding my writing into AI, I can also analyze my own style, identify recurring themes, and even spot patterns in my thinking that I wasn't consciously aware of. It's like having a conversation with your past self, discovering insights you'd forgotten you had.
The lesson here is simple but powerful: don't treat your published content as write-once, forget-forever material. Every piece you create is a potential building block for future work. Even if the platform you're using doesn't make this easy, AI tools can help you reclaim and repurpose your own intellectual output.
Whether you're writing newsletters, blog posts, or even social media content, think of each piece as contributing to your personal content library. With the right tools and a bit of AI assistance, your past self becomes one of your most valuable sources for your personal AI assistants.
Your writing isn't just content; it is your externalized thinking, ready to be rediscovered and built upon. Start treating it that way.
📰 AI Update
LM Studio: Now Free for Professional Use [LM Studio blog]
LM Studio is a user-friendly tool that allows you to run open-source language models locally on your machines. So far it was only free for private use. In a recent blog article, LM Studio has announced that they remove the need for a commercial license for workplace use.
Kyutai TTS: Open-Weight Voice AI That Outperforms ElevenLabs [Kyutai website]
Kyutai, the French AI research lab, has released an impressive new text-to-speech (TTS) model as open-weights that solves a major limitation of current voice synthesis tools. While most TTS systems require the complete text before they can start generating audio, Kyutai TTS can process text as it's being written, making it perfect for real-time applications where an AI assistant speaks while still thinking. With just 220 ms latency the 1.6 billion parameter model outperforms competitors like ElevenLabs in accuracy benchmarks.
Besides the TTS model with hundreds of voices but no voice cloning, they also release Unmute as open-source. The latter turns text LLMs into a voice AI, and on their demo page, unmute.sh, you can test this with the open-weight model Mistral-Small-3.2.
🌟 Media Recommendation
Podcast: Simon Sinek on Finding Meaning When Life Feels Overwhelming
In a world where AI is rapidly transforming how we work and communicate, Simon Sinek's recent appearance on the Modern Wisdom podcast offers timeless wisdom about what remains uniquely human: our need for genuine connection and purpose.
Sinek tackles a phenomenon many of us recognize but haven't fully articulated: how modern work has become our primary source of purpose, community, and even social life. As traditional sources of meaning like religious communities and local organizations have declined, we've placed enormous pressure on our jobs to fulfill needs they were never designed to meet. This shift helps explain why people are increasingly quitting when work doesn't align with their values or provide a sense of belonging.
Key insights that resonated with me:
Redefining Friendship: Sinek offers a simple but powerful definition - friendship is "agreeing to grow together." This framework applies whether we're talking about personal relationships, romantic partnerships, or even professional collaborations. It's about mutual support and intentional development.
The Power of Small Actions: Rather than grand gestures, lasting relationships are built through consistent, small acts of kindness - saying good morning before checking your phone, offering someone a drink, or simply being present. These "innocuous" actions accumulate into something profound over time.
Emotional Intelligence in Practice: When someone is upset, facts won't help, as you need to meet emotions with emotions. This principle becomes increasingly important as we navigate a world where AI can provide all the facts we need, but human empathy remains irreplaceable.
My take: As AI handles more of our analytical tasks, the skills Sinek discusses (empathy, authentic relationship-building, and creating genuine human connection) become our most valuable assets.
This episode serves as a perfect reminder that while AI can enhance our capabilities, it's our relationships and sense of community that ultimately give life meaning. Worth listening to while taking a walk or during your commute, as the insights will stick with you long after the episode ends.
Disclaimer: This newsletter is written with the aid of AI. I use AI as an assistant to generate and optimize the text. However, the amount of AI used varies depending on the topic and the content. I always curate and edit the text myself to ensure quality and accuracy. The opinions and views expressed in this newsletter are my own and do not necessarily reflect those of the sources or the AI models.