Feed aggregator
New York Seizes a Dozen Celebrity Deepfake Websites
Tech billionaires aren't juvenile science-fiction nerds, they just can't read
Article URL: https://www.avclub.com/billionaires-lack-media-literacy-thiel-musk-zuckerberg
Comments URL: https://news.ycombinator.com/item?id=49699767
Points: 1
# Comments: 0
When Everyone Has the Perfect Resume
Article URL: https://relocateme.substack.com/p/what-im-seeing-in-the-software-engineer
Comments URL: https://news.ycombinator.com/item?id=49699758
Points: 1
# Comments: 0
HuggingFace/tau: A Python port of Pi's minimalist coding agent
Article URL: https://github.com/huggingface/tau
Comments URL: https://news.ycombinator.com/item?id=49699751
Points: 1
# Comments: 0
AI Is in Dangerous Hands
Article URL: https://www.wheresyoured.at/ai-is-already-in-dangerous-hands/
Comments URL: https://news.ycombinator.com/item?id=49699738
Points: 1
# Comments: 0
Associations of computer gaming with incident dementia and cognitive functions (2024)
Article URL: https://link.springer.com/article/10.1186/s13195-024-01496-7
Comments URL: https://news.ycombinator.com/item?id=49699732
Points: 1
# Comments: 0
Wikivibes – Write the encyclopedia you wished existed
Article URL: https://wikivibes.org/
Comments URL: https://news.ycombinator.com/item?id=49699701
Points: 1
# Comments: 0
Fitting Neural Textures and PBR Material Maps with ES (No Backprop)
Article URL: http://richg42.blogspot.com/2026/09/fitting-neural-texture-and-pbr-material.html
Comments URL: https://news.ycombinator.com/item?id=49699696
Points: 1
# Comments: 0
Compact Neural Appearance Models for Efficient Gaussian Splatting
Article URL: https://fhahlbohm.github.io/efficient-gaussian-appearance/
Comments URL: https://news.ycombinator.com/item?id=49699682
Points: 1
# Comments: 0
Show HN: Cognition-Claude-proxy – Use Devin model catalog with Claude Code
Article URL: https://github.com/coderexpert123/cognition-claude-proxy
Comments URL: https://news.ycombinator.com/item?id=49699675
Points: 1
# Comments: 0
MazeDIY – free browser tool that turns any photo into a printable maze
Article URL: https://mazediy.com
Comments URL: https://news.ycombinator.com/item?id=49699672
Points: 1
# Comments: 0
Blizzard Revives 'StarCraft' Franchise with Open-World Sci-Fi Shooter
Article URL: https://hypebeast.com/2026/9/blizzard-open-world-starcraft-sci-fi-shooter-2030-release-info-teaser-revealed
Comments URL: https://news.ycombinator.com/item?id=49699671
Points: 1
# Comments: 0
Why don't machine learning research agents overfit?
Article URL: https://www.amazon.science/blog/why-dont-machine-learning-research-agents-overfit
Comments URL: https://news.ycombinator.com/item?id=49699648
Points: 1
# Comments: 0
We trained the fly to close B2B SaaS deals
Article URL: https://tryopine.com/flyslop
Comments URL: https://news.ycombinator.com/item?id=49699631
Points: 2
# Comments: 0
How is climate change affecting conifers and their ability to sequester carbon?
Article URL: https://www.nature.com/articles/d41586-026-02847-0
Comments URL: https://news.ycombinator.com/item?id=49699619
Points: 1
# Comments: 0
Sixteen Locks Ought to Be Enough for Anybody
Article URL: https://thebuild.com/blog/sixteen-locks-ought-to-be-enough-for-anybody/
Comments URL: https://news.ycombinator.com/item?id=49699611
Points: 1
# Comments: 0
Show HN: Rebuno - An open-source runtime for production agents
I have been building Rebuno, an open-source execution runtime for operating AI agents in production. It manages execution state and guardrails across agents built with different frameworks.
Agents run as HTTP services. Rebuno dispatches work through signed webhooks and keeps execution state in Postgres. Agents submit tool and model calls to Rebuno as steps before executing them.
For each new step, Rebuno evaluates per-agent YAML policies, independently of the model's prompt. It can allow the call, deny it, or hold it for human approval. The agent executes permitted calls and reports their outcomes. Policy decisions and step outcomes are both recorded in one append-only event log.
After an interruption, the agent starts its handler from the top. Steps with recorded outcomes return those outcomes instead of executing again. A crash can also leave a tool with no recorded outcome. Tools marked safe_to_retry run again, and tools marked at_most_once fail as indeterminate. How the agent handles an indeterminate step is up to its implementation.
There are Python and TypeScript SDKs, and the repo includes examples using LangChain, CrewAI, Pydantic AI, Vercel AI SDK, and Mastra.
Rebuno is free to self-host and released under the MIT licensed.
I'd be interested in hearing how others are running or operating agents in production.
Blog post with diagrams: https://rebuno.io/blog/introducing-rebuno
Getting started: https://docs.rebuno.io/getting-started
Comments URL: https://news.ycombinator.com/item?id=49699591
Points: 1
# Comments: 0
When LLM judges agree, should we believe them?
Article URL: https://www.amazon.science/blog/when-llm-judges-agree-should-we-believe-them
Comments URL: https://news.ycombinator.com/item?id=49699590
Points: 2
# Comments: 0
Show HN: DNSMint – Give a sandbox its own domain and HTTPS cert in one API call
Article URL: https://dnsmint.com/
Comments URL: https://news.ycombinator.com/item?id=49699572
Points: 1
# Comments: 0
