fp.
newest
Open in hackernews
LearnHarness – An Agent Harness for Learning
https://github.com/amosblomqvist/learnharness
3
•
abhaynayar
•
1h ago
Volvo Car Sinks to Record Low as CEO Says China Market Gone
https://www.bloomberg.com/news/articles/2026-10-23/volvo-car-deepen-job-cuts-china-sales-slump
1
•
toomuchtodo
•
33s ago
•
0 comments
Learning Is Now Optional
https://www.ericekholm.com/posts/learning-is-now-optional/
1
•
ekholm_e
•
1m ago
•
0 comments
A status-line monkey that gets more anxious the longer you code
https://pypi.org/project/jitters-cli/
1
•
vibeoverflow
•
1m ago
•
0 comments
How I teach with an iPad
https://theopenpresenter.com/blog/how-i-teach-with-an-ipad
1
•
michaelsalim
•
2m ago
•
0 comments
AmiJa, a cycle-exact Amiga emulator written in Java
https://www.reddit.com/r/amiga/comments/1xftc5s/amija_amiga_emulator_in_java/
1
•
doener
•
2m ago
•
0 comments
Find Creators to Play Your Game
https://marketyour.games/
1
•
prophittcorey
•
2m ago
•
1 comments
Show HN: Small agents that do one job, in Go
https://tomte.run/
1
•
berkaycubuk
•
4m ago
•
0 comments
Ask HN: What is your experience with keeping data secure
1
•
akimkern
•
4m ago
•
0 comments
AI Beneficiaries: Investing in Second-Order Effects – Update: 100x AI Cultures [pdf]
https://www.morganstanley.com/content/dam/im/assets/publication/thought-leadership/article/articl...
1
•
bluepeter
•
4m ago
•
0 comments
Malayalam: A University Course
https://ml.matj.io
1
•
mathewjacob
•
5m ago
•
1 comments
Reverse engineering a Russian ML driven drone detector
https://github.com/KonradIT/bandera-sa
1
•
konraditurbe
•
5m ago
•
0 comments
Anthropic Claude Models Contain AI Text Watermarks: What Does This Mean for You?
https://www.oreilly.com/radar/the-latest-anthropic-claude-models-contain-ai-text-watermarks-what-...
1
•
gmays
•
6m ago
•
0 comments
Pushing Bluetooth Out to 12 Kilometers
https://hackaday.com/2026/10/24/pushing-bluetooth-out-to-12-kilometers/
1
•
josephscott
•
7m ago
•
0 comments
Winners and Losers If the Data Center Buildout Is Delayed
https://www.apollo.com/wealth/insights-news/insights/daily-spark/winners-and-losers-if-the-data-c...
1
•
RickJWagner
•
7m ago
•
0 comments
NYPD in negotiations to use new AI-driven technology to track phones
https://www.nydailynews.com/2026/10/24/nypd-in-negotiations-to-use-new-ai-driven-technology-will-...
1
•
slowin
•
7m ago
•
0 comments
Fili – scriptable Zadig alternative for replacing USB drivers on Windows
https://github.com/PyLuke73/fili
1
•
PyLuke73
•
7m ago
•
0 comments
What'sUP Choiceland
https://wucland.substack.com/p/wuc-october-26-2026
1
•
jgudmuns
•
7m ago
•
0 comments
Torrents Through Mullvad on FreeBSD
https://www.frimin.fr/posts/torrents-through-mullvad/
1
•
gearnode
•
8m ago
•
0 comments
YouTube AV2 streams will be available for Web as soon as early 2027
http://aomedia.org/press%20releases/AOMedia-Rapid-AV2-Adoption-Across-the-Video-Ecosystem/
1
•
HelloUsername
•
9m ago
•
0 comments
KibernetykMono – A Geometric Monospace Typeface
https://pavlokilko.github.io/KibernetykMono/
1
•
thw_9a83c
•
9m ago
•
0 comments
The Inherent Hierarchy of Money (2012) [pdf]
https://sites.bu.edu/perry/files/2019/04/Mehrling_P_FESeminar_Sp12-02.pdf
1
•
mooreds
•
10m ago
•
0 comments
Show HN: Arc-decide – an MIT 0.4B decision model for RAG checks that runs on CPU
https://docs.docsgpt.cloud/Models/decision-model
1
•
alextttty
•
10m ago
•
0 comments
Anthropic's IPO filing shows explosive growth, risks and staggering costs
https://www.reuters.com/legal/transactional/anthropic-opens-books-ais-explosive-growth-staggering...
2
•
tartoran
•
11m ago
•
0 comments
Anthropic S-1: Inside the 335-page IPO plan
https://www.ft.com/content/4b2b9c0e-adfb-4362-800f-950f5b905984
1
•
Betelbuddy
•
11m ago
•
0 comments
R/open source just banned AI, non-English speakers, and more
https://www.theverge.com/ai-artificial-intelligence/1018217/reddit-opensource-ai-ban
1
•
isaacfrond
•
12m ago
•
0 comments
Mini: The Minimal Language
https://minilanguage.com/
3
•
surprisetalk
•
14m ago
•
0 comments
Show HN: DeepSeek v4.1 Flash-1017 Safety Evaluation
https://twitter.com/FangcunLeap/status/2114719692137136245
1
•
mayziem
•
15m ago
•
0 comments
Vegemite is not a human right
https://www.cnn.com/2026/10/26/australia/prisoner-vegemite-rejected-judge-scli-intl
4
•
sumnole
•
16m ago
•
4 comments
Engineering Evolved. The Rest Didn't. That's Why Your AI ROI Sucks
https://kerno.io/blog/why-your-ai-roi-sucks/
1
•
karimtr
•
17m ago
•
0 comments
Show HN: ToiletBench - Can AI find the toilet without a sign?
https://behu-kea.github.io/toilet-bench/
2
•
RedGreenBlack
•
19m ago
•
0 comments
Load More