Alvin

A guy used a Kindle for 4 years before he realized he was using it wrong. He read 60+ books on it. Highlighted hundreds of passages. Never adjusted a single setting beyond font size. His sister-in-law a librarian who's read 800+ books on her Kindle sat next to him on a flight and watched him read for 20 minutes. She finally said: "Can I show you something? You're missing the 9 features that make this thing actually useful. Amazon hides them 4 menus deep. Every Kindle owner I know reads way slower because of it." She changed 9 settings in 6 minutes. He finished his next book in half the usual time. Remembered twice as much. Looked up zero words on his phone. Here's everything she showed him 🧵

5. Noise Cancellation with one AirPod. Most earbuds disable ANC the moment you remove one bud. Apple has engineered the AirPods Pro 2 differently, allowing ANC use with one bud. You get some noise neutralization rather than none when in single earbud mode. [Geeky Gadgets](geeky-gadgets.com/mac-productivi…) Settings → Accessibility → Noise Cancellation with One AirPod → ON. For anyone who keeps one earbud out to stay partially aware at a desk, in a meeting, walking this is the setting that makes it actually.

4. Conversation Awareness auto-pause when you speak. You're listening to music at your desk. A colleague walks up and starts talking. Conversation Awareness automatically detects when you start speaking and lowers media volume while enhancing voices around you. It helps maintain natural conversations without removing your AirPods. [Hawkdive](hawkdive.com/hidden-macos-s…) Settings → your AirPods → Conversation Awareness → ON. No more yanking an earbud out. No more awkward "hold on" gestures. The AirPods detect you talking and step aside. Music comes back the moment the conversation ends.

A guy used AirPods Pro 2 for 2 years. He pressed play. He pressed pause. He answered calls. He switched to noise cancellation on planes. That was it. 4 features. On a $249 device with 30+. His coworker a sound engineer at a recording studio borrowed them for 10 minutes at lunch and handed them back with a look. "You've never opened Settings once, have you? Apple built 9 features into these that change how you work, sleep, hear, travel, and take calls. They're buried 3 menus deep and never mentioned in any ad. You've been wearing the world's most underused productivity tool." He spent 11 minutes in Settings that evening. His calls got clearer. His focus got deeper. He hasn't touched his phone during a workout since. Here's every feature his coworker showed him 🧵

Most people use NotebookLM the wrong way. They only ask for summaries — and end up with average results. Here are 10 advanced NotebookLM prompts that help you learn faster, think deeper, and truly understand your sources. 🔖 Save this for later.

A working mom of 3 spent 12 hours a week on household errands. Grocery runs. Drugstore stops. Target for diapers. Costco for paper goods. Birthday gift shopping. School supply trips. Hardware store for the random light bulb she forgot. She had Amazon Prime for 7 years. She used it like a regular shopping app. Search, add to cart, check out. Repeat. Her sister-in-law a project manager who runs her entire household from Amazon in under 30 minutes a month sat next to her and watched her shop. She finally said: "You're working 10x harder than you need to. Amazon built 8 features that turn this app into a household operating system. You've never opened any of them. Let me show you." 12 minutes of setup. 11 hours a week back. Here's everything she set up 🧵

I handed CLAUDE my SALARY. It told me I was 6 months from never needing one again. These are the 7 prompts behind my exit plan:

@ExhibitionsGM Yeah👍

@Alvin1492840 Have today that for me, these are all reasons to read a real book most of the time… I can see it being useful for study…

A guy used a Kindle for 4 years before he realized he was using it wrong. He read 60+ books on it. Highlighted hundreds of passages. Never adjusted a single setting beyond font size. His sister-in-law a librarian who's read 800+ books on her Kindle sat next to him on a flight and watched him read for 20 minutes. She finally said: "Can I show you something? You're missing the 9 features that make this thing actually useful. Amazon hides them 4 menus deep. Every Kindle owner I know reads way slower because of it." She changed 9 settings in 6 minutes. He finished his next book in half the usual time. Remembered twice as much. Looked up zero words on his phone. Here's everything she showed him 🧵

A guy used Audible for 6 years and never opened the Settings menu. 300+ hours of listening. He'd just press play and live with whatever the app gave him. Narrators sounded muffled. Volume was always too low in the car. His phone died every time he downloaded a new book. He blamed his headphones. He blamed his car stereo. He almost canceled the subscription. His coworker a sound engineer at a recording studio borrowed his phone on a flight and watched him fumble with the volume slider for 3 minutes. He finally said: "Can I see your settings? I'm guessing you've never touched them. Audible buries the good stuff. 4 minutes and I'll fix your entire listening experience." He fixed 8 things. The guy hasn't adjusted his volume in 4 months. Books sound clearer. Downloads finished 3x faster. He sleeps better. Here's exactly what the sound engineer changed 🧵

Andrej Karpathy stopped using AI to write code. The co-founder of OpenAI. The man who built Tesla's Autopilot vision team from scratch. The person who coined the term "vibe coding." In April 2026, he announced that a large fraction of his LLM token budget was no longer going into manipulating code, it was going into manipulating knowledge. Then he published a single markdown file on GitHub explaining what he had built instead. It got 17 million views. 13,000 GitHub stars. Dozens of community implementations within a week. He called it the LLM Wiki. And the idea behind it is so simple it is almost embarrassing that nobody published it sooner. Here is the problem it solves. Most people's experience with LLMs and documents looks like RAG you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works. But the LLM is rediscovering knowledge from scratch on every question. There is no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM works this way. ChatGPT file uploads work this way. Most RAG systems work this way. Every session starts from zero. The AI never learns the territory. It just searches it again. Karpathy's pattern is the opposite. Instead of retrieving from raw documents every time, the LLM builds and maintains a persistent, structured wiki — and answers questions from the compiled knowledge rather than the raw fragments. Here is how the architecture works. Three layers. **Layer 1 — Raw sources.** Your curated documents. Articles, papers, PDFs, meeting notes, screenshots. These are immutable — the LLM reads them but never modifies them. This is your source of truth. The moment you start editing raw files by hand, you have two systems of record and no way to tell which one is true. **Layer 2 — The wiki.** A directory of LLM-generated markdown files. Summaries, entity pages, concept pages, comparisons, a master index, a chronological log. The LLM owns this layer entirely. It creates pages, updates them when new sources arrive, maintains cross-references, and keeps everything consistent. You read it. The LLM writes it. **Layer 3 — The schema.** A CLAUDE.md or AGENTS.md file that tells the LLM how the wiki is structured, what conventions to follow, and what workflows to run. This is the config that turns a generic chatbot into a disciplined wiki maintainer. Karpathy's phrase captures the whole thing: "Obsidian is the IDE. The LLM is the programmer. The wiki is the codebase." Here is what happens when you drop a new source into the system. You save an article into raw/. You tell the LLM to ingest it. The LLM reads the source, writes a summary page, updates the master index, creates or updates entity pages for every person, company, or concept mentioned, creates or updates concept pages for every idea, adds cross-references between related pages, and logs the ingest in the activity record. A single source might touch 10 to 15 wiki pages. This is the bookkeeping that humans abandon — filing, cross-referencing, updating related entries, noting contradictions. The exact work that kills every personal knowledge system you have ever started. The LLM does it tirelessly. Every time. Without forgetting. Here is the key distinction from RAG. RAG re-derives an answer from raw chunks on every query and accumulates nothing. The LLM Wiki compiles sources into structured, linked pages once — and questions are answered from that built artifact. The analogy: raw/ is source code, wiki/ is the compiled executable. Knowledge that is compiled is retrieved. Knowledge that is not is rediscovered from scratch. And here is the rule Karpathy emphasizes most. Lint the knowledge. Treat the wiki like code and run health checks. Ask the model to find contradictions between pages, surface low-confidence claims, list orphan pages, and flag entities that drifted into two spellings. A contradiction is information — it means two sources disagree and now you know where to look. Skipping the lint is how a wiki quietly rots while the graph still looks impressive. Start small. Begin with ten sources, not ten thousand. Get ingest, query, and lint to feel natural before you add complexity. The first few ingests will be messy. Naming conventions will change. That is normal. A small wiki you actually use beats a beautiful architecture you abandon in week three. The community response tells you how much this resonated. Within a week of Karpathy's gist, the community produced dozens of implementations full Python agents, Obsidian integrations, wiki compilers, web interfaces. The pattern works with Claude Code, Codex, OpenCode, Gemini CLI, and any LLM agent that can read and write files. You do not need any of them. The entire system works with nothing but an LLM agent and a file system. Paste the pattern into your CLAUDE.md and Claude Code becomes your wiki maintainer. Here is why this matters more than another AI tool. Every personal knowledge system you have ever tried Notion, Evernote, Roam, Obsidian, died the same way. Not because the tool was bad. Because the maintenance was unsustainable. The filing. The tagging. The cross-referencing. The updating when new information arrived. The bookkeeping that makes a knowledge base useful is the exact work nobody wants to do. Karpathy's insight is that the bookkeeping is exactly what LLMs are good at. Tirelessly reading, summarizing, filing, linking, updating, and maintaining consistency — without getting bored, without forgetting, without deciding it is too tedious and abandoning the project in week four. You curate sources and ask questions. The LLM does the bookkeeping. The wiki compounds over time every source you add and every question you ask makes it richer. The tedious part of maintaining a knowledge base is not the reading or the thinking. It is the bookkeeping. And the bookkeeping just got automated. Source: Andrej Karpathy · GitHub Gist · AI Builder Club · Vanja. io · MindStudio · April 2026 ( Link in the comments)

@liamaimans Amazing share

If this helped, two asks: 1. Repost the first post so the next listener stops fighting their app. 2. Follow @Alvin1492840 I break down the hidden settings, features, and tactics that companies bank on you not knowing. Next thread: the 7 audiobooks under 4 hours that change how you think about money.

The uncomfortable truth: Audible's default settings aren't built for the listener. They're built for the lowest common denominator. Quieter audio. Slower playback. Mid-range muffled. 30-second skip-backs that punish your attention. No equalizer because it's been requested for 8+ years and never built. The sound engineer's last line: "You weren't a bad listener. You were using the app the way the company set it up to be used. That's not the same as the way it should be used." 4 minutes of settings changes. 6 years of degraded listening, undone.