0. Goodbye Claude, hello Pi (caleb.software)
728 points · 448 comments · by karagenit
A longtime Claude Code user says he is switching to Pi, citing Claude Code’s instability, excessive feature additions, IDE limitations, duplicated configuration formats, and system-prompt overhead, while praising Pi’s stability, minimalism, and customizability. [src]
1. Goldman Sachs is pulling Claude Code from its bankers and traders (thestateofai.com)
692 points · 384 comments · by TheWiggles
Goldman Sachs has restricted Claude Code and GitHub access to engineers and strats, cutting off bankers and traders amid productivity, security and compliance concerns, even as the bank remains a lead underwriter for Anthropic’s planned IPO. [src]
2. Nvidia CEO Jensen Huang's letter to staff about affair (nvidianews.nvidia.com)
491 points · 528 comments · by karimf
Nvidia CEO Jensen Huang addressed employees in a company letter amid reports about an alleged affair and related scrutiny. [src]
3. Nvidia Board of Directors Is Investigating Jensen Huang (nvidianews.nvidia.com)
454 points · 340 comments · by vincentchau
NVIDIA’s board is launching an independent investigation into reports that CEO Jensen Huang was targeted by a foreign intelligence effort, while saying it has no reason to believe national security or company secrets were compromised. [src]
4. Ask HN: What are you working on? (November 2026)
180 points · 605 comments · by david927
A November 2026 Hacker News thread invites users to share what they’re working on and what topics have recently sparked their curiosity. [src]
5. Outlook.com is rejecting email from domains without DMARC (mailroute.net)
334 points · 252 comments · by xmddmx
Since November 5, 2026, Outlook.com has reportedly rejected some messages from domains without DMARC records—including those from small senders—making a `v=DMARC1; p=none` TXT record the recommended fix. [src]
6. Llama-modchip: 284B DeepSeek on 64GB RAM, no GPU (github.com)
402 points · 180 comments · by joshcsimmons
GitHub project `llama-modchip` claims patched CPU-first inference can run a quantized 284B DeepSeek-V4-Flash model in under 59 GiB of RAM at up to 14.7 tokens per second without a GPU, alongside faster 30B MoE models. [src]
7. Person of Interest (krthr.co)
297 points · 188 comments · by krthr
The author reflects on revisiting *Person of Interest*, finding its surveillance themes increasingly relevant while nostalgically recalling early-2010s internet culture, tech communities, SOPA/PIPA, Megaupload, and the show’s enduring quality. [src]
8. Tesla Hasn't Had a Successful New Product Launch Since JB Straubel Stepped Down (cleantechnica.com)
198 points · 261 comments · by gumby
CleanTechnica argues that Tesla has not launched a significant, successful new product since CTO JB Straubel stepped down in 2019, citing the Cybertruck’s poor performance and questioning whether Straubel’s departure contributed to the company’s stalled product development. [src]
9. KetchUp – Open-source SketchUp alternative built with Rust and WebGPU (ketchup3d.github.io)
356 points · 94 comments · by dcoulmbe
BenchUp is a free, open-source Rust-based 3D modeling app for woodworking and furniture design, offering browser and desktop versions with precise measurements, local autosave, SketchUp imports, and OBJ exports. [src]
Brought to you by ALCAZAR. Protect what matters.