Feed aggregator
Why the Internet Built for Friends Felt More Meaningful Than AI
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
Show HN: Deliberate practice applied to coding interviews
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
Khal – CLI Calendar App
Article URL: https://github.com/pimutils/khal
Comments URL: https://news.ycombinator.com/item?id=47300447
Points: 3
# Comments: 0
Video Shows US Tomahawk Missile Strike Next to Girls' School in Iran
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
Reasoning models struggle to control their chains of thought, and that's good
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
MemryX MX3M.2 AI Accelerator Module [pdf]
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
The Misconception That Air Supremacy Has Been Achieved over Iran
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
The Biggest Identity Sandpiles and How to Compute Them
Article URL: https://eavan.blog/posts/big-identity-sandpiles.html
Comments URL: https://news.ycombinator.com/item?id=47300343
Points: 2
# Comments: 2
AI Disruption: Doomsday Scenario or Fizzle? [video]
Article URL: https://www.youtube.com/watch?v=TbOAtQU89eA
Comments URL: https://news.ycombinator.com/item?id=47300338
Points: 1
# Comments: 0
Ask HN: Please restrict new accounts from posting
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
Liberation and Immanence
Article URL: https://www.gojiberries.io/liberation-and-immanence/
Comments URL: https://news.ycombinator.com/item?id=47300076
Points: 1
# Comments: 0
AI CEOs worry the government will nationalize AI
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
C++ is essentially secular Judaism
Article URL: https://twitter.com/norpadon/status/2030561676643180832
Comments URL: https://news.ycombinator.com/item?id=47300069
Points: 1
# Comments: 0
Warning: Chuwi CoreBook X uses AMD Ryzen 5 5500U instead of 7430U
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
Telehamster, a landline network to connect in an old school screen-free way
Article URL: https://www.youtube.com/watch?v=UhVi3smmvTs
Comments URL: https://news.ycombinator.com/item?id=47300053
Points: 1
# Comments: 0
OpenAI might end up on the right side of history
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
Website Testing on Steroids
Article URL: https://www.grepular.com/Website_Testing_on_Steroids
Comments URL: https://news.ycombinator.com/item?id=47300036
Points: 1
# Comments: 0
Show HN: FileMayor – A zero-dependency, local-first file organization engine
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
Show HN: I mapped the exact minute every Pixar movie makes you cry
Article URL: https://sheets.works/data-viz/pixar-cry-chart
Comments URL: https://news.ycombinator.com/item?id=47299997
Points: 1
# Comments: 0
