newest
Open in hackernews
The SEP has achieved what Wikipedia can only dream of (2015)
https://qz.com/480741/this-free-online-encyclopedia-has-achieved-what-wikipedia-can-only-dream-of
8
•
Tomte
•
2h ago
Do you need a Vector Search Database?
https://fin.ai/research/do-you-really-need-a-vector-search-database/
1
•
thunderbong
•
1m ago
•
0 comments
Show HN: Reactylon – The React Framework for XR
https://www.reactylon.com/docs
1
•
simonedev_
•
1m ago
•
0 comments
My Coding Game Makes $1M per Month [video]
https://www.youtube.com/watch?v=tMkpiFIW8Xg
1
•
simonpure
•
1m ago
•
0 comments
How to Survive Until the Singularity
https://singularitysurvivalguide.com/
1
•
yakhinvadim
•
1m ago
•
0 comments
Show HN: The first stage of my indie game, made with my own engine, is ready
https://reprobateonline.xyz/
1
•
delduca
•
4m ago
•
0 comments
Ask HN: Have you used AI to apply for jobs for you? How did it go?
1
•
givemeethekeys
•
5m ago
•
0 comments
ArduPilot – open-source autopilot system
https://ardupilot.org/
1
•
danielovichdk
•
5m ago
•
0 comments
Show HN: The Wikle – Improve quality of AI answers with human source curation
https://www.thewikle.com/w/Welcome_to_THE_WIKLE!
1
•
systemstops
•
5m ago
•
0 comments
Things You Can Do with CSS Today (2021)
https://www.smashingmagazine.com/2021/02/things-you-can-do-with-css-today/
1
•
Tomte
•
6m ago
•
0 comments
Libogc Allegations Rock Wii Homebrew Community
https://hackaday.com/2025/05/01/libogc-allegations-rock-wii-homebrew-community/
1
•
Tomte
•
7m ago
•
0 comments
China will likely match U.S. AI model capabilities this year
https://blog.heim.xyz/chinas-ai-models-are-closing-the-gap/
2
•
speckx
•
7m ago
•
0 comments
A New, Simpler License for Max and Mojo
https://www.modular.com/blog/a-new-simpler-license-for-max-and-mojo
1
•
roflcopter69
•
8m ago
•
0 comments
Important 2025 Plex Updates
https://www.plex.tv/blog/important-2025-plex-updates/
1
•
YourCupOTea
•
8m ago
•
0 comments
BD+05 4868 Ab is a planet rapidly disintegrating, producing a comet-like tail
https://news.mit.edu/2025/astronomers-discover-planet-thats-rapidly-disintegrating-producing-comet-tail-0422
1
•
gmays
•
8m ago
•
0 comments
When You Get to Be Smart Writing a Macro
https://tonsky.me/blog/hashp/
1
•
stopachka
•
10m ago
•
0 comments
Looking at my time data from 2012 to 2025
https://sachachua.com/blog/2025/04/looking-at-my-time-data-from-2012-to-2025/
3
•
lr0
•
11m ago
•
0 comments
Epic vs. Apple
https://twitter.com/dhh/status/1917834147076153348
3
•
kashifr
•
11m ago
•
0 comments
Faxed Arabic prescriptions: a medication error waiting to happen? (2012)
https://pubmed.ncbi.nlm.nih.gov/22178294/
2
•
Bluestein
•
13m ago
•
0 comments
VOYO – Multilingual emotional voice agents for India
https://www.voyo.club/
2
•
prabhsingh888
•
14m ago
•
1 comments
New Study: Waymo is reducing serious crashes and making streets safer
https://waymo.com/blog/2025/05/waymo-making-streets-safer-for-vru
2
•
prossercj
•
14m ago
•
1 comments
Harrods latest retailer to be hit by cyber attack
https://www.bbc.com/news/articles/c62x4zxe418o
1
•
gnabgib
•
16m ago
•
0 comments
Show HN: Spellbound – open-source Chrome and Safari AI writing extensions
https://github.com/stanlee000/spellbound/tree/main/safari-extension
1
•
stanlee_o
•
16m ago
•
0 comments
UK Retail Under Siege: Cyber Attacks, Supply Chains and Lessons
https://blog.edned.net/uk-retail-cyber-attacks/
1
•
_edned
•
18m ago
•
0 comments
RAG can make AI models riskier and less reliable, new research shows
https://www.zdnet.com/article/rag-can-make-ai-models-riskier-and-less-reliable-new-research-shows/
1
•
CrankyBear
•
19m ago
•
0 comments
Ai2's new small AI model outperforms similarly-sized models from Google, Meta
https://techcrunch.com/2025/05/01/ai2s-new-small-ai-model-outperforms-similarly-sized-models-from-google-meta/
1
•
tntpreneur
•
19m ago
•
0 comments
London's National Gallery purchases a painting by an unknown artist for $20M
https://www.theartnewspaper.com/2025/05/01/national-gallery-buys-mysterious-altarpiece-for-20m
2
•
giuliomagnifico
•
19m ago
•
0 comments
10 years of Earning to Give (2023)
https://forum.effectivealtruism.org/posts/gxppfWhx7ta2fkF3R/10-years-of-earning-to-give
1
•
eamag
•
20m ago
•
0 comments
Twelve theses and predictions on "AGI" (falsely so-called)
https://programnotes.blog/2025/04/05/twelve-theses-and-predictions-on.html
1
•
wyclif
•
21m ago
•
0 comments
People who don't ask me questions drive me crazy. Why are they like that?
https://www.theguardian.com/wellness/2025/may/01/why-dont-people-ask-questions-in-conversation
3
•
creer
•
24m ago
•
0 comments
Howich of Open AI does Y Combinator own?
1
•
investAI
•
24m ago
•
0 comments