fp.

Project Itohs Harmony and the under explored extremes of alignment theory

1•calmkeepai•1m ago•0 comments

Upgrading in Site or Relocation? Impacts of Slum Renewal Policies

https://www.nber.org/papers/w34560

1•felineflock•2m ago•0 comments

Running 9 AI agents and realizing we needed a control plane

https://www.runshift.ai/

2•devincrane•4m ago•2 comments

Why Private School Isn't Worth the Cost

https://ofdollarsanddata.com/why-private-school-isnt-worth-the-cost/

1•amin•4m ago•0 comments

Federal Cyber Experts Called Microsoft's Cloud "A Pile of Shit", yet Approved It

https://www.propublica.org/article/microsoft-cloud-fedramp-cybersecurity-government

1•hn_acker•5m ago•1 comments

Product has always been the bottleneck, it's just clearer now

https://kwakubiney.github.io/posts/Product-Has-Always-Been-The-Bottleneck-Its-Just-Clearer-Now/

1•kwakubiney•5m ago•0 comments

Tennis XGBoost autoresearch loop hacked probability scoring on plateau anxiety

https://www.nickoak.com/posts/tennis-xgboost-autoresearch/

1•buildoak•5m ago•1 comments

Not just energy: How the Iran war could trigger a global food crisis

https://www.aljazeera.com/economy/2026/3/18/not-just-energy-how-the-iran-war-could-trigger-a-glob...

1•akbarnama•7m ago•0 comments

Show HN: Ibkr-CLI – A local-first Interactive Brokers CLI built for AI agents

https://github.com/fatwang2/ibkr-cli

1•fatwang2•7m ago•1 comments

Budget Security, SoC 2 compliant pentests fast, cheap, self service

https://budgetsecurity.com/en

1•BudSec•8m ago•0 comments

Meta vowed to stop illegal financial ads. It failed 1k times in a week

https://www.reuters.com/sustainability/boards-policy-regulation/meta-vowed-stop-illegal-financial...

2•1vuio0pswjnm7•8m ago•0 comments

Veto: Permission policy engine and LLM firewall for AI coding agents

https://www.vetoapp.io/

1•damienhauser•8m ago•1 comments

All my clients wanted a carousel, now it's an AI chatbot

https://adele.pages.casa/md/blog/all-my-clients-wanted-a-carousel-now-it-s-an-ai-chatbot.md

1•speckx•9m ago•0 comments

U+237C ⍼ Right angle with downwards zigzag arrow

https://ionathan.ch/2022/04/09/angzarr.html

1•fooqux•9m ago•0 comments

Show HN: Studio – a drag-and-drop PDF workspace for multi-document workflows

https://localpdf.online/app/studio

1•ulinycoin•9m ago•1 comments

The internet ruined customer service. AI could save it

https://www.a16z.news/p/the-internet-ruined-customer-service

3•7777777phil•10m ago•0 comments

World ID wants to put a unique human identity on every AI agent

https://arstechnica.com/ai/2026/03/world-id-wants-you-to-put-a-cryptographically-unique-human-ide...

2•Brajeshwar•12m ago•0 comments

Researchers uncover iPhone spyware capable of penetrating millions of devices

https://www.reuters.com/technology/researchers-uncover-iphone-spyware-capable-penetrating-million...

2•1vuio0pswjnm7•12m ago•0 comments

Your Job Is Dead. Autopsies for the AI Transition

https://www.job.rip/

2•bixvolt•13m ago•0 comments

React SSR Framework Showdown: TanStack, React Router and Next.js Under Load

https://blog.platformatic.dev/react-ssr-framework-benchmark-tanstack-start-react-router-nextjs

1•TheAnkurTyagi•13m ago•0 comments

Heaps do lie: debugging a memory leak in vLLM

https://mistral.ai/news/debugging-memory-leak-in-vllm

1•shiftingleft•15m ago•0 comments

Where did the lightweight web go?

https://medium.com/@hmnt.bazooka/where-did-the-lightweight-web-go-1c1ee2caa447

3•hmntyee•15m ago•0 comments

Show HN: WeFrames One shared photo a day from Apple Shared Albums

1•sillendev•15m ago•0 comments

Show HN: KonbiniAPI – One API for Instagram and TikTok Data, Same JSON Shape

https://konbiniapi.com

1•FdezRomero•16m ago•0 comments

A ngrok-style secure tunnel server written in Rust and Open Source

https://github.com/joaoh82/rustunnel

2•joaoh82•18m ago•1 comments

Anthropic's Hidden Vercel Competitor "Antspace"

https://aprilnea.me/en/blog/reverse-engineering-claude-code-antspace

1•AprilNEA•18m ago•0 comments

Why Good Economic Policies Can Fail – The Need for Incentives and Reminders

https://www.nominalnews.com/p/why-good-economic-policies-can-fail

1•NomNew•19m ago•0 comments

Show HN: Applemail – Cursor skill for full macOS Mail.app control

https://github.com/Don-Yin/apple-mail

1•DonYin•19m ago•0 comments

A drop-in binary replacement to migrate from MinIO to RustFS

https://github.com/rustfs/rustfs/issues/2212

1•KevinatRustFS•19m ago•1 comments

Belgium launches secure messaging app for government staff

https://www.belganewsagency.eu/belgium-launches-secure-messaging-app-for-government-staff

3•whynotmaybe•20m ago•0 comments

Open in hackernews

Inference Price Index – what AI inference costs across 8 providers

https://www.volthq.dev/

1•JackArnot•1h ago

Comments

JackArnot•1h ago

https://open.substack.com/pub/touchgrasscapital/p/inference-...

Some of the numbers that surprised me:

Llama 70B: 4.4x price spread across 5 providers for the same weights

DeepSeek R1 output: $2/M tokens on Hyperbolic vs $60/M on OpenAI o1 (30x)

GPT-5.4 mini launched yesterday at $0.25/$2.00 — same price as GPT-5 mini, 2x faster

I track 8 providers in real time. The MCP server is free and runs inside Claude and Cursor:

npx volthq-mcp-server --setup

Publishing weekly. Would love feedback on what data would be most useful to track next.