fp.

Show HN: Offline 180.9M-parameter LLM and Agent inference on ESP32-P4

https://github.com/cyfrit/p-for-llm

1•Cyfrit•6s ago•0 comments

Foundation Model to Predict Airspace

https://archer.com/tech-blog/zee-predicting-the-airspace

1•polzounov•10s ago•0 comments

Zero-Day to Zero Doubt: AI-Powered CVE Forensics in an Afternoon

https://blog.quent.in/blog/2026/07/31/zero-day-to-zero-doubt-ai-powered-cve-forensics-in-an-after...

1•sylvainkalache•28s ago•0 comments

Celeris-1: A diffusion LLM benchmarked at 2,082 output tokens/s

https://artificialanalysis.ai/models/celeris-1

1•jn2clark•1m ago•0 comments

Situational Awareness hedge fund meltdown was a warning shot BofA CEO says

https://www.cnbc.com/2026/08/05/bofa-brian-moynihan-situational-awareness-meltdown-was-a-warning-...

2•mapping365•2m ago•0 comments

Qwen 3.0 Image Pro

https://www.qwencloud.com/models/qwen-image-3.0-pro

1•theanonymousone•3m ago•0 comments

SpaceX revenue rockets while AI spending burns billions

https://www.theregister.com/ai-and-ml/2026/08/05/spacex-revenue-rockets-while-ai-spending-burns-b...

2•speckx•4m ago•0 comments

Token-Native Storage: Read and Write in Your Agent's Language

https://arxiv.org/abs/2608.02376

1•eln1•5m ago•0 comments

Microsoft Tells Engineers 'Tokenmaxxing Is Not What We Are Optimizing For'

https://www.404media.co/microsoft-tells-engineers-tokenmaxxing-is-not-what-we-are-optimizing-for/

2•Brajeshwar•5m ago•0 comments

The State of AI in the Binocular Viewer Industry

https://www.adamarice.com/ai-binocular-viewers/

1•adamarice•7m ago•0 comments

Mixture-of-Kittens: Cursor's open-source MoE megakernel for NVL72s

https://cursor.com/blog/mixture-of-kittens

1•alexhandru•7m ago•0 comments

Anthropic Inks $10B Computing Deal with New Startup Volta Park

https://www.bloomberg.com/news/articles/2026-08-04/anthropic-inks-10-billion-computing-deal-with-...

2•htrp•8m ago•0 comments

Trump administration drafting ban on Chinese fiber transceivers

https://www.reuters.com/world/trump-administration-drafting-ban-chinese-data-center-devices-sourc...

2•HotGarbage•9m ago•0 comments

SpaceX shares sink after first earnings report reveals AI spending plans

https://www.bbc.com/news/articles/c0qvpveg20vo

2•Lyngbakr•11m ago•0 comments

Amazon.com vs Perplexity AI [pdf]

https://cdn.ca9.uscourts.gov/datastore/opinions/2026/08/04/26-1444.pdf

2•miohtama•11m ago•0 comments

The FANUC industrial empire [video]

https://www.youtube.com/watch?v=8iSSPDLRlC0

1•ripe•12m ago•0 comments

East Africa's iconic snow peaks will soon disappear for good

https://theconversation.com/east-africas-iconic-snow-peaks-will-soon-disappear-for-good-what-will...

1•speckx•14m ago•0 comments

Proxmox VE now available for ARM64

https://forum.proxmox.com/threads/proxmox-virtual-environment-now-available-for-64-bit-arm-arm64....

2•geerlingguy•14m ago•0 comments

The "Disability Dongle": Why Silicon Valley Hates Me and You

https://sightlessscribbles.com/disability-dongle/

1•calcifer•18m ago•1 comments

Show HN: Twocal – Calendar sync that doesn't silently drop events

https://twocal.app/

1•gsudarkoff•19m ago•0 comments

Show HN: On This Minute: 525,600 message slots, one per minute of the year

https://onthisminute.com

1•circadex•19m ago•0 comments

Show HN: Test the **** Out of Your Containers

https://github.com/bxrne/dstest

2•bxrne•20m ago•0 comments

Show HN: WeaveMark – a specification language for reusable prompts

https://github.com/paulosalem/weavemark

1•paulosalem•20m ago•1 comments

Show HN: I asked an LLM what it would say to God, then made it a website

https://sekor.eu.org/x/i-am-god/

1•modinfo•21m ago•1 comments

Why a Raspberry Pi shouldn't be powered through its GPIO pins

https://82mhz.net/posts/2026/08/why-a-raspberry-pi-shouldn-t-be-powered-through-it-s-gpio-pins/

1•speckx•22m ago•1 comments

Show HN: Formatif – A local-first, offline media converter

https://formatif.net

1•hadikaraki•22m ago•0 comments

Amazon VP: Stack Ranking and Pips, Working with Bezos, His Promotions

https://www.youtube.com/watch?v=40-ENZmqcz0

1•johnbarron•23m ago•1 comments

Show HN: Language learning using speech-analyzed native speaker video

https://reelang.com

1•stuess•24m ago•0 comments

Show HN: Colophon:Show HN: Colophon: Google Docs Writing Process Log

https://colophon.firl.nl/demo

1•normanbell•26m ago•0 comments

We analyzed 144 hours of creator recordings. A quarter of it was silence

https://vidclean.net/blog/how-much-of-a-recording-is-silence/

2•TheBuciyo•27m ago•1 comments