frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Tabsmith-lint catches Chrome Web Store rejections before you submit

https://github.com/rsub122/tabsmith-lint
1•rsub122•1m ago•0 comments

Launchpad Is Down

https://launchpad.net/
2•rebelwebmaster•1m ago•1 comments

As SuperAgers age, they make at least twice as many new neurons as their peers

https://news.northwestern.edu/stories/2026/02/as-superagers-age-they-make-at-least-twice-as-many-...
1•Teever•1m ago•0 comments

Layoffs hit Bellevue-based video game studio behind 'Destiny' franchise

https://www.seattletimes.com/business/local-business/layoffs-hit-bellevue-based-video-game-studio...
2•ynac•8m ago•0 comments

Staff Framework – Tornando Desejos Em Metas Verificáveis (Smart Method )

https://fidelissauro.dev/staff-smart-methods/
1•lsferreira42•10m ago•0 comments

New $1M Vesuvius Challenge Grand Prize will be announced in coming days

https://www.youtube.com/watch?v=96oTlQm0KBw
1•janpaul123•13m ago•0 comments

Anthropic Accuses Alibaba of Largest AI Distillation Attack: 28.8M Fraudulent

https://yipzap.com/anthropic-accuses-alibaba-of-largest-ai-distillation-attack-28-8m-fraudulent-e...
2•noida•15m ago•0 comments

The Ticks That Cause Red-Meat Allergies Are Spreading Across the U.S.

https://www.wsj.com/health/the-ticks-that-cause-red-meat-allergies-are-on-the-move-this-summer-b5...
2•rawgabbit•16m ago•1 comments

Razor‐Sharp Edge–The Yakutat Slab Dissecting South‐Central Alaska

https://pubs.geoscienceworld.org/ssa/tsr/article/6/2/230/731510/Razor-Sharp-Edge-The-Yakutat-Slab...
2•gnabgib•17m ago•0 comments

Radxa Orion O6N Review: The Powerful and Silent ARM64 Beast

https://boilingsteam.com/radxa-orion-o6n-review/
1•ekianjo•19m ago•0 comments

How agents are transforming work

https://openai.com/index/how-agents-are-transforming-work/
1•gmays•23m ago•0 comments

An Audit of the Bible and the God in It

https://medium.com/freedomofthought/an-audit-of-the-bible-and-the-god-in-it-0c425c3a2194
1•raynchad•24m ago•0 comments

Ask HN: How do you know when an Ad campaign is doing well?

1•TheRickyRed•27m ago•4 comments

Ask HN: Has anyone deployed AI tools in a trade or field service

1•VaderAi•27m ago•1 comments

Chasing Likes, Losing Connection: Youth Mental Health in the Digital Era

https://medium.com/freedomofthought/chasing-likes-losing-connection-youth-mental-health-in-the-di...
1•raynchad•27m ago•0 comments

Linux Foundation Launches Akrites to Defend FOSS from AI-Enabled Exploits

https://www.phoronix.com/news/Akrites
2•LorenDB•30m ago•0 comments

The Windows Phone that "killed" Nokia

https://www.instagram.com/reel/DZvCjtLM-UQ/
2•cromka•34m ago•1 comments

Sony Confirms 'Significant' Bungie Layoffs

https://www.ign.com/articles/bungie-announces-layoffs-as-it-admits-destiny-2-fell-short-of-expect...
5•mikhael•34m ago•1 comments

DropItDown – Drop a file, get Markdown your AI agent can read (macOS)

https://dropitdown.novashang.com
1•novashang•40m ago•0 comments

How do you get good ideas for startups?

https://www.paulgraham.com/ideas.html
1•Alien1Being•44m ago•1 comments

Extropic is Rethinking Computing

https://extropic.ai/writing/thermodynamic-computing-from-zero-to-one
1•binyu•46m ago•0 comments

CS2-10k: A Large-Scale Egocentric Counter-Strike 2 Dataset

https://reka.ai/news/cs2-10k-a-large-scale-egocentric-counter-strike-2-dataset
1•ilreb•49m ago•0 comments

Using a Rust macro for concise newtypes

https://adenalhardan.com/#rust-macros-concise-newtypes
1•adenalhardan•51m ago•0 comments

Ask HN: How did you set up a multi-agent orchestration for personal use?

2•cromka•57m ago•3 comments

A Deep Dive on China's "LineShine" All-CPU, Exaflops-Class Supercomputer

https://www.nextplatform.com/hpc/2026/06/25/a-deep-dive-on-chinas-lineshine-all-cpu-exaflops-clas...
1•jonbaer•1h ago•0 comments

Anthropic's philosopher answers your questions [video]

https://www.youtube.com/watch?v=I9aGC6Ui3eE
2•binyu•1h ago•1 comments

Silicon Valley Has an Empathy Vacuum (2016)

https://www.newyorker.com/business/currency/silicon-valley-has-an-empathy-vacuum
20•rmason•1h ago•12 comments

Van Halen test

https://en.wikipedia.org/wiki/Van_Halen_test
7•teleforce•1h ago•3 comments

Show HN: FastPlay, a fast minimal Windows video player built in Rust

https://github.com/CalvinSturm/FastPlay
2•CalvinSturm•1h ago•0 comments

Investors bet on AI again after Micron reports 346% sales jump

https://www.cnn.com/2026/06/25/business/micron-results-ai-stocks-volatility
4•dabinat•1h ago•1 comments
Open in hackernews

Show HN: DeepSeek Flash inverted the economics of agent products

https://www.rtrvr.ai/blog/code-as-plan-deepseek-flash-text-only-browser-agent
8•arjunchint•2h ago
There is an adversarial relationship between developers and big model labs.

Model labs charged developers higher API prices to subsidize their own agent harness offerings. Think Anthropic charging 5x higher Claude API prices to subsidize consumer subscriptions. So Cursor in a way was subsidizing their own direct competitor.

DeepSeek V4 Flash totally inverted this relationship. Now you have a model that beats even Sonnet in some benchmarks and is totally opensourced. Now inference providers are racing to the bottom to optimize and give cheaper hosting. Every player with a non-SOTA is now racing to swap over to stop paying the big model lab tax, even Microsoft is switching Copilot to use DeepSeek.

On switching over to Deepseek:

- we noticed over a 100x cost decrease while similar or better performance then Gemini 3 Flash

- insane saving from the cached input tokens: $0.002/1 Million tokens

- both DeepSeek Flash and GLM 5.2 are text-only models, so clearly multimodal training is not worth the additional cost. Language is just a much more efficient sparse representation of the world/reasoning than vision

- we had a early bet on a text-only web agent harness, and now with DeepSeek this results in unique cost advantages.

- we rewrote our harness as a callable DSL library that a model can generate code to execute on. DeepSeek has proven phenomenal on code generation to drive an agent harness.

- I would highly recommend everyone to rewrite their harness to be text-only and callable via executable code leveraging DeepSeek V4 Flash.