fp.

Mixture of Experts (Moe): How Transformers Scale Without Activating Everything

https://chizkidd.github.io//2026/08/10/mixture-of-experts/

1•ibobev•26s ago•0 comments

A Noob Learns FFT

https://entropicthoughts.com/fft

1•ibobev•1m ago•0 comments

How Stalled models, missed deadlines, and burnout led to unraveling of DeepMind

https://fortune.com/2026/08/10/how-stalled-models-missed-deadlines-and-staff-burnout-lead-to-the-...

1•theanonymousone•1m ago•0 comments

911 calls answered by AI in New Orleans [video]

https://www.youtube.com/watch?v=psllRvOotwI

1•soraminazuki•1m ago•0 comments

Show HN: Turn any topic into a 60-second 3-host AI podcast (no signup)

https://demades.ai/try

1•samiam4238•1m ago•0 comments

Tritium Nuclear Battery Offers Low-Risk Space Energy

https://spectrum.ieee.org/tritium-nuclear-battery

1•rbanffy•1m ago•0 comments

FBI investigating North Korean remote IT staffer working for US agency

https://federalnewsnetwork.com/cybersecurity/2026/08/fbi-investigating-north-korean-remote-it-sta...

1•jaredwiener•1m ago•0 comments

Software Engineering Is Cool, but Is It Interesting?

https://ertu.dev/posts/5_software_engineering_is_cool_but_is_it_interesting/

1•theanonymousone•2m ago•0 comments

Tuscolo2026h2 write-path degradation on August 9th

https://groups.google.com/a/chromium.org/d/msgid/ct-policy/b1b5e8a2-b010-4e2a-b8c8-91c1d1b7ca91%4...

1•l2dy•2m ago•0 comments

$580M undersea cable rerouted to avoid the grave of Dobby the House Elf

https://www.tomshardware.com/networking/usd580-million-undersea-cable-rerouted-to-avoid-the-grave...

1•rbanffy•4m ago•0 comments

What Louis CK would think about Nvidia's circular financing

https://wrongquestions.substack.com/p/what-louis-ck-would-think-about-nvidias

1•yamirghofran•4m ago•0 comments

Want people to use cheaper AI models? Make expensive models slower

https://enterpilot.io/blog/make-expensive-ai-models-slower/

1•santiago-pl•5m ago•1 comments

Pizza Box Project Stack

https://cblgh.org/posts/2026-07-31-pizza-box-project-stack/

1•surprisetalk•6m ago•0 comments

How much would you pay for a smarter baby?

https://www.economist.com/business/2026/08/10/how-much-would-you-pay-for-a-smarter-baby

1•AsyncBanana•8m ago•0 comments

Firefox stunts on Microsoft neutering adblockers: 'uBO isn't going anywhere'

https://www.pcgamer.com/software/browsers/firefox-stunts-on-edge-as-microsoft-gears-up-to-neuter-...

2•HelloUsername•8m ago•0 comments

What survived rebuilding our SRE investigation agent six times

https://engineeringexec.tech/posts/evolution-of-an-sre-agent/

1•mbleterman•8m ago•0 comments

Layer Scope: How I used $20 of compute to make a new way to look at LLMs

https://blog.lwarfield.dev/layer-scope/

1•lwarfield•9m ago•1 comments

I wrote a C inference engine that runs BitNet 2B at 15–18 tok/s on a Pi 5

https://github.com/geisten/geistlib

1•geisten•9m ago•1 comments

Show HN: Afk, a CLI that scaffolds unattended cron and bash and Telegram agents

https://github.com/nahuelsoria/afk-agents

1•nahuelsoria•10m ago•0 comments

A Great Deal of Information About Hacker News

https://vale.rocks/posts/hacker-news

3•OuterVale•11m ago•0 comments

Wing Commander 1 source reconstruction

https://github.com/neuromancer/wc1-re

1•galapago•11m ago•0 comments

Show HN: C# Game Engine with its own scripting language and IDE

https://github.com/ArcadeMakerSources/ArcadeMaker

1•am-gm•11m ago•0 comments

Should You Self-Host LLM Inference? Cost and Risk Guide

https://theaiengineer.substack.com/p/should-you-self-host-inference

1•speckx•12m ago•0 comments

Scientists Release Biggest 2D Map of the Universe

https://newscenter.lbl.gov/2026/08/10/scientists-release-biggest-2d-map-of-the-universe/

2•NKosmatos•12m ago•1 comments

Chinese-Made Components on UK Naval Drone Cameras Sent Info to Beijing

https://www.kyivpost.com/post/82061

2•Teever•13m ago•0 comments

Show HN: Coordinate coding agents and Git worktrees around features

https://github.com/jan-bogaerts/md2

2•jbo76•13m ago•1 comments

I generate launch videos from a URL and never render one on a server

https://venturehand.com

1•hristoff•15m ago•0 comments

LLMs empowered agent-based modeling and simulation (2024)

https://www.nature.com/articles/s41599-024-03611-3

1•hamburgererror•15m ago•0 comments

Show HN: Write.md, a free, open-source, themeable Markdown editor for macOS

https://writemd.app/

2•danielbilekq•16m ago•0 comments

Show HN: A new chat interface for Ollama

https://github.com/bctnry/OpenSyzygy

3•kasumispencer2•16m ago•0 comments

Mixture of Experts (Moe): How Transformers Scale Without Activating Everything

A Noob Learns FFT

How Stalled models, missed deadlines, and burnout led to unraveling of DeepMind

911 calls answered by AI in New Orleans [video]

Show HN: Turn any topic into a 60-second 3-host AI podcast (no signup)

Tritium Nuclear Battery Offers Low-Risk Space Energy

FBI investigating North Korean remote IT staffer working for US agency

Software Engineering Is Cool, but Is It Interesting?

Tuscolo2026h2 write-path degradation on August 9th

$580M undersea cable rerouted to avoid the grave of Dobby the House Elf

What Louis CK would think about Nvidia's circular financing

Want people to use cheaper AI models? Make expensive models slower

Pizza Box Project Stack

How much would you pay for a smarter baby?

Firefox stunts on Microsoft neutering adblockers: 'uBO isn't going anywhere'

What survived rebuilding our SRE investigation agent six times

Layer Scope: How I used $20 of compute to make a new way to look at LLMs

I wrote a C inference engine that runs BitNet 2B at 15–18 tok/s on a Pi 5

Show HN: Afk, a CLI that scaffolds unattended cron and bash and Telegram agents

A Great Deal of Information About Hacker News

Wing Commander 1 source reconstruction

Show HN: C# Game Engine with its own scripting language and IDE

Should You Self-Host LLM Inference? Cost and Risk Guide

Scientists Release Biggest 2D Map of the Universe

Chinese-Made Components on UK Naval Drone Cameras Sent Info to Beijing

Show HN: Coordinate coding agents and Git worktrees around features

I generate launch videos from a URL and never render one on a server

LLMs empowered agent-based modeling and simulation (2024)

Show HN: Write.md, a free, open-source, themeable Markdown editor for macOS

Show HN: A new chat interface for Ollama

Show HN: Amber – An open-source C99 Q/K array engine for quant finance

Mixture of Experts (Moe): How Transformers Scale Without Activating Everything

A Noob Learns FFT

How Stalled models, missed deadlines, and burnout led to unraveling of DeepMind

911 calls answered by AI in New Orleans [video]

Show HN: Turn any topic into a 60-second 3-host AI podcast (no signup)

Tritium Nuclear Battery Offers Low-Risk Space Energy

FBI investigating North Korean remote IT staffer working for US agency

Software Engineering Is Cool, but Is It Interesting?

Tuscolo2026h2 write-path degradation on August 9th

$580M undersea cable rerouted to avoid the grave of Dobby the House Elf

What Louis CK would think about Nvidia's circular financing

Want people to use cheaper AI models? Make expensive models slower

Pizza Box Project Stack

How much would you pay for a smarter baby?

Firefox stunts on Microsoft neutering adblockers: 'uBO isn't going anywhere'

What survived rebuilding our SRE investigation agent six times

Layer Scope: How I used $20 of compute to make a new way to look at LLMs

I wrote a C inference engine that runs BitNet 2B at 15–18 tok/s on a Pi 5

Show HN: Afk, a CLI that scaffolds unattended cron and bash and Telegram agents

A Great Deal of Information About Hacker News

Wing Commander 1 source reconstruction

Show HN: C# Game Engine with its own scripting language and IDE

Should You Self-Host LLM Inference? Cost and Risk Guide

Scientists Release Biggest 2D Map of the Universe

Chinese-Made Components on UK Naval Drone Cameras Sent Info to Beijing

Show HN: Coordinate coding agents and Git worktrees around features

I generate launch videos from a URL and never render one on a server

LLMs empowered agent-based modeling and simulation (2024)

Show HN: Write.md, a free, open-source, themeable Markdown editor for macOS

Show HN: A new chat interface for Ollama