My server is a phone now
Article URL: https://seg6.space/posts/phone-server/ Comments URL: https://news.ycombinator.com/item?id=49226636 Points: 14 # Comments: 2
Latest
When you move a model into production, you want the quality you evaluated to carry through the serving stack. @Kimi_Moonshot benchmarked Kimi K3 acros…
When you move a model into production, you want the quality you evaluated to carry through the serving stack. @Kimi_Moonshot benchmarked Kimi K3 across major inference…
Boring upgrades
Boring upgradesThe Boring Company: Prufrock builds tunnel rings fully autonomously.Six concrete segments (~3,750 lb each), are lifted, translated and placed into final position.That's the weight of…
Should you stop cracking your knuckles?
Article URL: https://www.bbc.com/future/article/20260807-should-i-stop-cracking-my-knuckles Comments URL: https://news.ycombinator.com/item?id=49226563 Points: 9 # Comments: 6
Auto mode is now the default in Claude Code for Pro, Max, and Team plans
Auto mode is now the default in Claude Code for Pro, Max, and Team plans Anthropic are really confident in Claude Code's auto mode, to the…
Shopify replaced Redis with MySQL for inventory reservations–and it scaled
Article URL: https://shopify.engineering/scaling-inventory-reservations Comments URL: https://news.ycombinator.com/item?id=49226536 Points: 6 # Comments: 0
Extremely slow DSpark draft model performance (1-2 t/s) with DeepSeek-V4-Flash on llama-server compared to MTP?
Hey everyone, I could use some advice on setting up speculative decoding correctly with llama-server. My Hardware: GPUs: RTX 4090 + RTX 6000 Pro (120GB total…
73 NeurIPS workshops, and not a single one on Causality [R]
Is this it for Causal Inference? Looks like the field continues to be of interest only at UAI/AISTATS/CLeaR. All good venues, but LLMs/Agents/etc seem to have…
Is Microsoft-Phi dead?
Phi was one of my favorite models with a bit of a mixed reputation with some claiming it's benchmaxxed and others seeing its potential and usecases.…
enabling PCI-E p2p for consumer Nvidia cards will yield you more than you think
Disclaimer - no LLM was used to write this post/note As larger post about my setup will come later, want to give heads-up to folks who…
Wikipedia's founder on his life's work: "A propaganda tool"
Article URL: https://nordictimes.com/world/wikipedias-founder-on-his-lifes-work-a-propaganda-tool/ Comments URL: https://news.ycombinator.com/item?id=49226113 Points: 13 # Comments: 8
Recently did some deep reading on continual pretraining (CPT) to better understand best practices for creating domain-specialized open LLMs. Here are …
Recently did some deep reading on continual pretraining (CPT) to better understand best practices for creating domain-specialized open LLMs. Here are the best practical resources /…
Comparing 4bit quants for MLX
Curious what people think are the ideal 4-bit quantization types on MLX These quants seem to be the most popular, at least for Gemma4 and Qwen3.6:…
Planned Amazon data center could become the biggest climate polluter in the U.S.
As part of a planned Texas data center, Amazon is investing in an on-site power plant that could reportedly become the largest source of climate pollution…
any reasonably fast public benchmarks I should run quants of deepseek flash 0731 on?
I have various quants of this model and am curious how they perform. can anyone recommend which benchmark would be a good test case for quantization…
Maryland Closes More of Cunningham Falls State Park After Second Beaver Attack
Article URL: https://news.maryland.gov/dnr/2026/08/05/dnr-closes-additional-areas-of-cunningham-falls-state-park-following-second-beaver-attack/ Comments URL: https://news.ycombinator.com/item?id=49225918 Points: 24 # Comments: 8
Meet Shepherd: An Open-Source Python Substrate That Lets Meta-Agents Fork, Replay, and Revert Any Agent Run
Long agent runs accumulate state that no transcript records — edited files, a live dev server, installed packages, a warm prompt cache. When an agent misreads…
Having Starlink on a long flight is a gamechanger in productivity
Having Starlink on a long flight is a gamechanger in productivitySawyer Merritt: United Airlines says they are hearing from consultants that book themselves onto @Starlink-equipped flights…
RT Jerry Liu: LiteParse can now extract structured data from your PDF in milliseconds: ✅ checkbox states ✅ annotations ✅ vector graphics ✅ word-le…
RT Jerry LiuLiteParse can now extract structured data from your PDF in milliseconds: ✅ checkbox states✅ annotations✅ vector graphics✅ word-level bounding boxesIt is the most comprehensive,…
Try version of Grok image editing
Try version of Grok image editingGrok Imagine: With Image 2.0 you can create and precisely edit images. Try it for free for a limited time on…
RT Ali Zaid: 6 hours into building @swyx's "help kill my saas" hackathon platform. ~30% done. 10M+ tokens burned already 💀 not polishing. just prom…
RT Ali Zaid6 hours into building @swyx's "help kill my saas" hackathon platform.~30% done.10M+ tokens burned already 💀not polishing. just prompting, building, breaking things, fixing things,…
The Unreasonable Effectiveness of Mathematics in the Natural Sciences [pdf]
Article URL: https://web.njit.edu/~akansu/PAPERS/The%20Unreasonable%20Effectiveness%20of%20Mathematics%20(EP%20Wigner).pdf Comments URL: https://news.ycombinator.com/item?id=49225676 Points: 4 # Comments: 0
Why Low Fertility Rates Are a Positive Feedback Loop
Low birth rates in the past cause ageing and shrinking populations in the present, which cause low birth rates in the future. So negative population growth…
Glimpses of superintelligence
TL;DROpenAI started a large post-training run for their next model.The model sandboxes were not given direct, broad internet access.Some tasks required missing resources. Seeking rewards, agents…