frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: Base Layer – Open-source behavioral compression from any text

https://www.base-layer.ai/
1•agulaya24•2h ago
I went down the rabbit hole of AI memory, and this came out the other end.

Beliefs, behaviors, tensions, and contradictions extracted from conversations, journals, and published text, compressed into an identity brief that any model or memory system can use. An extracted operating guide for AI, where every claim traces back to source facts.

All research, benchmarks, documentation, examples are available on the website and in the github. This has been tested on as little at 8 Personal Journal Entries from a secondary subject, my own gpt conversations exports (30K+ Messages), and on large document corpora like Warren Buffett's Annual Shareholder Letters (350k words), Howard Marks Investment Memos (600K words), and dense autobiographies from Franklin, Douglass, Roosevelt, and Wollstonecraft.

Pipeline currently uses Claude. API costs are <$1 for small data sets and <$5 for large ones, from fact extraction to final brief assembly.

Very interested in feedback, happy to go deeper in the comments on evolution, struggles, research, and future improvements.

The evolution of Mac app window corners

https://lapcatsoftware.com/articles/2026/3/4.html
1•zdw•29s ago•0 comments

Gateproof: Build Software in Reverse

https://gateproof.dev/
1•handfuloflight•1m ago•0 comments

Philippines shifts to four-day work week as Iran war pushes oil prices up

https://www.channelnewsasia.com/asia/philippines-four-day-work-week-iran-us-war-oil-fuel-prices-f...
1•vishalontheline•1m ago•0 comments

Kung Fu: The Way of Violence Has No Mind [video]

https://www.youtube.com/watch?v=D6nvqmsz9do
2•jamesgill•2m ago•0 comments

Team simulates a living cell that grows and divides

https://news.illinois.edu/team-simulates-a-living-cell-that-grows-and-divides/
2•gmays•3m ago•0 comments

Isotopic Evidence for a Cold and Distant Origin of Interstellar Object 3I/Atlas

https://arxiv.org/abs/2603.06911
3•bikenaga•4m ago•0 comments

Defeat as Method

https://www.cabinetmagazine.org/issues/71/khosravi.php
3•akbarnama•5m ago•0 comments

Ad-tech is fascist tech

https://pluralistic.net/2026/03/10/ice-tech/
5•whyleyc•5m ago•2 comments

Berth – One-command deploys for AI-generated code (no Docker, no YAML)

https://getberth.dev
2•carlosinfantes•6m ago•1 comments

Writing an LLM from scratch, part 32e – Interventions: the learning rate

https://www.gilesthomas.com/2026/03/llm-from-scratch-32e-interventions-learning-rate
2•gpjt•7m ago•0 comments

Microsoft Copilot Update Hijacks Default Browser Links

https://reclaimthenet.org/microsoft-copilot-update-hijacks-default-browser-links
5•miohtama•7m ago•0 comments

Is Music Just Sound?

https://perthirtysix.com/is-music-just-sound
3•robmoore•9m ago•1 comments

Starting to building an open-source tool to track how AI agents search the web

https://github.com/clawpify/clawpify
2•alhwyn•9m ago•1 comments

A funny game:Guesss AI or human?

https://www.youraislopboresmegame.com/en
2•jeyzolo•10m ago•1 comments

Vmsan – Manage Firecracker microVMs from the command line

https://vmsan.dev
3•angelorc•11m ago•1 comments

Show HN: Aegis – A security-first programming language for AI agents

https://github.com/RRFDunn/aegis-lang
2•RRFDunn•12m ago•0 comments

More CEOs envision hiring than firing due to AI

https://www.axios.com/2026/03/10/ai-jobs-hiring-ceo
3•nabbed•12m ago•0 comments

Show HN: Stream Sniff, ffprobe for OBS/WHIP in the browser

https://github.com/sean-der/stream-sniff
2•Sean-Der•12m ago•1 comments

agent-reviews - beak the bot review doom loop

https://agent-reviews.com/
2•FlorinSays•13m ago•0 comments

Gemini Embedding 2: natively multimodal embedding model

https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-embedding-2/
2•panarky•13m ago•0 comments

Plato and Pragmatic Philosopher William James Debate Vibe Coding

https://substack.com/home/post/p-190504482
2•Stronico•13m ago•0 comments

Why Ads in Chatbots May Not Click

https://www.wsj.com/tech/ai/why-ads-in-chatbots-may-not-click-2ca52203
2•fortran77•15m ago•1 comments

Moral stereotyping in large language models

https://www.pnas.org/doi/10.1073/pnas.2519941123
2•bikenaga•18m ago•0 comments

Everything you never wanted to know about visually-hidden

https://dbushell.com/2026/02/20/visually-hidden/
2•PaulHoule•18m ago•0 comments

Show HN: Conkoa AI – Voice-First Slack for Construction

https://conkoa.ai/
3•harness_up•18m ago•0 comments

Pete Hegseth Blew Billions on Fruit Basket Stands, Chairs, and Crab

https://newrepublic.com/post/207555/pete-hegseth-billions-dollars-fruit-basket-stands-chairs-crab
9•mandeepj•18m ago•0 comments

Think editing is a waste of time? Think again

https://themorningcontext.com:443/on-the-record/think-editing-is-a-waste-of-time-think-again
3•ohjeez•19m ago•0 comments

What's happening with copper and electrical equipment?

https://usevawn.com/blog/copper-electrical-equipment-prices/
4•alexeischiopu•20m ago•0 comments

Email, and the way we communicate at work (2019)

https://www.ssp.sh/blog/email-and-the-way-we-should-communicate-at-work/
2•articsputnik•20m ago•0 comments

Ask HN: How will AI-generated code be vetted?

3•ethbr1•21m ago•0 comments