Rereading a quick set of notes I jotted down from Brian Chesky’s Figma Config ’23 talk. I find myself coming back to his definition of true simplification, which applies to every craft—from coding to art
The more elaborate your system for capturing ideas, the less time you spend actually thinking with them.
It's like constantly taking photos—you document the experience instead of having it.
Relying on LLMs for writing and thinking may erode critical thinking, create dependency on external knowledge sources, and undermine intellectual autonomy. Study: https://t.co/wTREKFQWuO [This post is actually written by an LLM]
1/3 Working on my personal projects, the concept of LOM started taking shape: Living Organized Memory. Or call it whatever, simply "the memory" (I just love how LOM sounds). Something that will probably become the default for many applications.
At night went to get the $100 tokens. A weekend to spend them all, I'm on my way. At a peak I pushed autonomous tasks to 5 VS Code instances and got that feeling: our bio context window is becoming the bottleneck. Commercial devs will be under more cognitive stress, not less
Free vibe engineering workshop this week!
Learn AI-assisted TypeScript development using open-source tools to code smarter, not harder.
✅ $100 in free API tokens
✅ Build a real TypeScript + React app
✅ Try the latest cutting edge models (Claude 4, Gemini 2.5 pro, GPT 4.1 and others)
✅ Zero cost. Zero catch.
Link below 👇
@neuotq Well, if a digital archive had two-way sync with our brains... o_O
But anyway, like any data storage, privacy and security is always a concern. Even our built-in human memory is not reliable while it rewrites itself during recall
3/3 This means apps like Obsidian or Logseq wouldn't require tedious curation and manual linking. I'm betting Dale Cooper's tape recorder in Twin Peaks will soon be more than a toy: voice input becomes something meaningful and connected. Scattered entries → LOM
2/3 It wasn't easy to build pre-AI but is now becoming tangible. I've been experimenting with SQL triggers, then exploring graph/RDF databases, though it feels like a rabbit hole for me personally (lacks exp). Now ready-made products are appearing. Yes, I mean mem0 and others.
Crazy.
Vector databases for AI memory just got replaced by... MP4 files!
You can now store millions of text chunks inside an MP4 file with Memvid, and still get lightning-fast semantic search! 🤯
No database needed. 100% Opensource!
repo below ↓
Anticipating the 'human first' trend. Like cloud and subscriptions aren't kings anymore. Things change even faster now. People will notice AI fatigue even before it gets our total default. I caught myself talking to a friend today the same way I chat with Claude, and that's bad
I loved saying: a good visual design fights entropy. AI codegen applies that more than ever to the dev's role. A new kind of entropy now — chaos quietly introduced by intelligent assistants. Treat AI like juniors: review carefully, even when it seems flawless. Especially then!
I've been exploring a zoomable timeline that makes it possible to see only the biggest events in your life, scaling to any level of granularity. It would need extra UI. But the magic of OLLOS is how naturally it reflects the changing importance of things https://t.co/NztfqaLY0x
Obenauer's OLLOS is a rich source of ideas I keep coming back to. The unified timeline is something I'm trying to approach in my own WIP, though in a slightly different way. What grabs me is allowing timeline entries to have varying degrees of importance https://t.co/X3x38NzGDS
Obenauer's OLLOS is a rich source of ideas I keep coming back to. The unified timeline is something I'm trying to approach in my own WIP, though in a slightly different way. What grabs me is allowing timeline entries to have varying degrees of importance https://t.co/X3x38NzGDS
At long last, I present to you: the OLLOS essay.
It's got everything: how it works & how I used it, design and concept iterations, future work for it and the OS, and what I learned about personal interfaces.
Enjoy: https://t.co/aMbL163tyK