Feed aggregator

At the beginning of the social internet there was a very simple promise. Technology would help people stay connected to those who were already part of their lives. There was no obsession with “audiences”, no creator economy, no algorithms deciding what you should watch next. The center of everything was human relationships. That is why platforms like Facebook, Orkut and even early YouTube felt meaningful to millions of people. They worked as extensions of real social life.

Today the situation is curious. We live in the most technologically advanced digital era in history. Artificial intelligence writes articles, generates images and can hold long conversations with us. Yet many people feel that these tools do not occupy the same emotional space in their lives that those early social platforms once did. To understand why, it helps to look carefully at how the social web evolved and what quietly changed along the way.

When Facebook launched in 2004 its purpose was extremely narrow. It was built to connect college students who already knew each other. Access required a university email address. There were no influencers, no viral video feeds, and no global audiences. People logged in to see photos from the weekend party, comment on a friend’s post, or check where classmates were studying abroad. The structure of the platform reflected this intention. The core of the experience was simply your list of friends.

That detail, which might seem trivial today, changed everything. Research from the Pew Research Center showed that the vast majority of Facebook connections were people users already knew in real life. On average only about 7 percent of connections were with people the user had never met offline. Most contacts came from school, family, work or existing social circles. https://www.pewresearch.org/internet/2011/06/16/social-networking-sites-and-our-lives-2/

Read the full content here:

Comments URL: https://news.ycombinator.com/item?id=47300512

Points: 2

# Comments: 0

I built InterviewTraner to fix how people prepare for coding interviews. Most people grind LeetCode problems randomly or follow static lists, which is not very effective. InterviewTraner uses a deliberate practice engine to intelligently schedule which problems you should work on and when.

How it works:

- 1,800+ LeetCode problems organized into a prerequisite graph across all major topics (arrays, trees, graphs, DP, etc.).

- The engine starts you on easy problems and only unlocks harder topics once you've demonstrated mastery of their prerequisites.

- Spaced repetition, mastery learning, interleaving, and other strategies ensure you practice what you need.

- Powerful filtering features to only practice a subset of problems when you are in a rush.

What makes it different from just doing LeetCode or some curated sets? The problems are static and you will have to waste time figuring out what to practice. InterviewTraner turns those static problems into a dynamic experience personalized and optimized for you.

I plan on eventually expanding the curriculum beyond coding interviews to anything in the software engineering process that can fit into short and repeatable exercises.

I originally built Trane, the practice engine powering it, to optimize my music practice. The core insight (skills have prerequisites and practice should respect that structure) applies equally well to coding interviews.

Pricing:

- Free lite license (arrays, strings, recursion, hash tables, stacks).

- $10/month for all 1,800+ problems (launch price with LAUNCH code, I plan the normal price to be $20/month).

Happy to answer questions about the deliberate practice approach, the scheduling algorithm, or anything else.

Links:

InterviewTraner: https://interviewtraner.com

Pictures Are For Babies (same engine applied to learning to read and write): https://picturesareforbabies.com

Trane: https://github.com/trane-project/trane

Comments URL: https://news.ycombinator.com/item?id=47300463

Points: 2

# Comments: 0

Article URL: https://github.com/pimutils/khal

Comments URL: https://news.ycombinator.com/item?id=47300447

Points: 3

# Comments: 0

Article URL: https://www.bellingcat.com/news/2026/03/08/video-shows-us-tomahawk-missile-strike-next-to-girls-school-in-iran/

Comments URL: https://news.ycombinator.com/item?id=47300440

Points: 6

# Comments: 1

Article URL: https://openai.com/index/reasoning-models-chain-of-thought-controllability/

Comments URL: https://news.ycombinator.com/item?id=47300432

Points: 3

# Comments: 1

Article URL: https://memryx.com/wp-content/uploads/2025/04/MX3-M.2-AI-Accelerator-Module-Product-brief-DEC25-Gold.pdf

Comments URL: https://news.ycombinator.com/item?id=47300429

Points: 2

# Comments: 0

Article URL: https://www.twz.com/air/unprecedented-new-view-of-blu-109-bunker-buster-strike-that-killed-hezbollahs-leadership

Comments URL: https://news.ycombinator.com/item?id=47300381

Points: 4

# Comments: 0

Article URL: https://eavan.blog/posts/big-identity-sandpiles.html

Comments URL: https://news.ycombinator.com/item?id=47300343

Points: 2

# Comments: 2

Article URL: https://www.youtube.com/watch?v=TbOAtQU89eA

Comments URL: https://news.ycombinator.com/item?id=47300338

Points: 1

# Comments: 0

I don’t know if I’m the only one, but I see lots of clearly AI generated posts recently in HN and mostly coming from new accounts (green), it is more noticeable in the Show HN section.

I wish the team can either restrict new accounts from posting or at least offer a default filtering where I can only see posts from accounts with certain criteria.

I don’t want to see HN becoming twitter, which is full of bots and noise, as this would be a really sad day.

Comments URL: https://news.ycombinator.com/item?id=47300329

Points: 68

# Comments: 43

Article URL: https://www.gojiberries.io/liberation-and-immanence/

Comments URL: https://news.ycombinator.com/item?id=47300076

Points: 1

# Comments: 0

Article URL: https://yro.slashdot.org/story/26/03/07/2058213/ai-ceos-worry-the-government-will-nationalize-ai

Comments URL: https://news.ycombinator.com/item?id=47300070

Points: 2

# Comments: 0

Article URL: https://twitter.com/norpadon/status/2030561676643180832

Comments URL: https://news.ycombinator.com/item?id=47300069

Points: 1

# Comments: 0

Article URL: https://www.notebookcheck.net/Warning-Chuwi-CoreBook-X-uses-AMD-Ryzen-5-5500U-instead-of-7430U.1242048.0.html

Comments URL: https://news.ycombinator.com/item?id=47300060

Points: 2

# Comments: 0

Article URL: https://www.youtube.com/watch?v=UhVi3smmvTs

Comments URL: https://news.ycombinator.com/item?id=47300053

Points: 1

# Comments: 0

note: I am in MENA, am not with the military in any way.

when i first read the statement by Dario, i was shocked by the fact the military was so dismissive about Ai safety (not to mention privacy). Seeing anthropic resist the military, I felt so proud of being a claude user to the point I deleted gpt right away. it's nice to see your fav products sync with your values.

but today, after thinking more about it, i realized something. for a government to allow one Ai company to dictate terms, it opens up a precedent for Ai companies in the future to resist governmental oversight. that might not be a big deal in 2020s, but in 2030s by all estimations many Ai companies will be big enough to resist entire governmental structures. Maybe not the US or China, but they will definitely be big enough not to be easily influenced.

those independent companies will eventually grow so large, no government can hope to tame them. i know that right now it seems impossible for a mere ccorp valued at less than a trillion to resist a government that spends 7 trillion each year. but zooming out, it feels likely that the next generation of Ai companies will be easily valued at 10T. if you look at a 2-year-old which just learned how to talk and suddenly he starts talking quantum, you can bet your a* he will grow up to be a powerhouse.

i know soft monetary power is very different than hard military power, but enough tokens of the first type can easily be converted into the second type if: 1. you have a sufficiently ambitious CEO. 2. the survival of the company is threatened in some way. I am not talking about AGI here, but good old private equity that does whatever it needs to survive. ruled by suits that have more loyalty to shareholders than anyone or anything else.

at the end of the day, corporations are ruled by dictators (they have to be), governments are not (not in the West at least). maybe just maybe we should NOT trust private equity to seek anything but profits. governments are manipulative and bloody, but at least we can vote.

Comments URL: https://news.ycombinator.com/item?id=47300047

Points: 1

# Comments: 0

Article URL: https://www.grepular.com/Website_Testing_on_Steroids

Comments URL: https://news.ycombinator.com/item?id=47300036

Points: 1

# Comments: 0

Hi HN,

My documents folder recently became an absolute disaster—thousands of unsorted PDFs, raw images, scattered .csv files, and code snippets from the last three years.

I looked for automated categorization tools, but every modern solution seemed to require uploading my local file metadata to a cloud UI or running an overly heavy background service. I explicitly wanted an offline, privacy-first engine.

So I built FileMayor. It’s a 100% local-first data organization engine built on Node.

A few technical properties I focused on:

Zero Runtime Dependencies: The core Node engine is pure. No vulnerabilities, no bloat, and minimal memory footprint. Deterministic Fallbacks: By default, files are sorted instantly into 12 hardcoded extensions/mime-type categories (Documents, Media, Archives, Code, etc.) using offline pattern matching. The Rollback Journal: Every file mutation is logged to a local .filemayor-journal.json. If an organization run ruins your directory structure, a single undo command reverses the entire batch operation instantly. AI SOPs (Opt-in): If you need complex directory trees, it parses .md or .txt Standard Operating Procedures, securely queries the Gemini API to transpile the intent into a locked YAML schema, and executes the file moves. I packaged versions for Windows, macOS (arm64), and Linux (.deb).

I'd really appreciate any feedback on the rollback journal architecture or the regex pattern matching approach!

Comments URL: https://news.ycombinator.com/item?id=47300013

Points: 1

# Comments: 0

Article URL: https://sheets.works/data-viz/pixar-cry-chart

Comments URL: https://news.ycombinator.com/item?id=47299997

Points: 1

# Comments: 0