Codex for almost everything

420 mikeevans 2026-04-16 17:26:21 UTC openai.com Llama Share 216

Codex now operates alongside developers, generating images, remembering preferences, and taking on repeatable work, with deeper support for developer workflows and native web integration. The update includes over 90 new plugins, improved automations, and memory features to enhance productivity and collaboration.

Claude Opus 4.7

1138 meetpateltech 2026-04-16 14:26:21 UTC anthropic.com Llama Share 849

Opus 4.7 is a notable improvement over Opus 4.6 with better vision, more tasteful and creative task completion, and improved cybersecurity capabilities. It handles complex tasks with rigor and consistency, and is available today across all Claude products and APIs.

PCI Express over Fiber [video]

74 mmastrac 2026-04-16 17:26:21 UTC youtube.com Embed Share 19

German Dog Commands

19 rolph 2026-04-16 19:26:21 UTC fluentu.com Llama Share 21

You're learning German and want to teach your dog new commands using Hundekommandos. This can help you learn and practice German words, especially verbs, and improve your grammar skills.

Cloudflare's AI Platform: an inference layer designed for agents

191 nikitoci 2026-04-16 13:26:21 UTC blog.cloudflare.com Llama Share 45

Cloudflare is launching AI Gateway, a unified inference layer that allows developers to access any AI model from any provider through one API. This simplifies AI development and management, enabling developers to build fast and reliable AI-powered applications.

TigerBeetle: A Trillion Transactions [video]

32 adityaathalye 2026-04-16 18:26:21 UTC youtube.com Embed Share 12

Qwen3.6-35B-A3B: Agentic coding power, now open to all

721 cmitsakis 2026-04-16 14:26:21 UTC qwen.ai Llama Share 339

Qwen3.6-35B-A3B is a sparse MoE model with 3B active parameters, delivering outstanding agentic coding performance and rivaling larger dense models. It's available as open weights on Hugging Face and ModelScope, and through Alibaba Cloud Model Studio API.

Qwen3.6-35B-A3B on my laptop drew me a better pelican than Claude Opus 4.7

99 simonw 2026-04-16 18:26:21 UTC simonwillison.net Llama Share 22

The author compares two AI models, Qwen3.6 and Claude Opus 4.7, on a pelican riding a bicycle benchmark. Qwen3.6 produces better results, making it a better choice for specific tasks.

Launch HN: Kampala (YC W26) – Reverse-Engineer Apps into APIs

49 alexblackwell_ 2026-04-16 15:26:21 UTC zatanna.ai Full Share 49

Zatanna maps human-only software workflows and exposes them as reliable APIs your AI agents and internal tools can call in production. No browser automation scripts to maintain.

Put your SSH keys in your TPM chip

67 type0 2026-04-16 15:26:21 UTC raymii.org Llama Share 69

User securely stores SSH private key in a hardware token or TPM for added security. They generate keys offline, back them up, and import them into the TPM for secure access to servers.

The future of everything is lies, I guess: Where do we go from here?

390 aphyr 2026-04-16 14:26:21 UTC aphyr.com Llama Share 400

The author is concerned about the impact of large language models on society, citing their potential to displace human workers and exacerbate existing problems. They advocate for slowing down the development of these models to give time for regulation and adaptation.

Circuit Transformations, Loop Fusion, and Inductive Proof

5 matt_d 2026-04-16 19:26:21 UTC natetyoung.github.io Prefix Share 0

During a visit to Cornell, Sam presented a bunch of hardware optimisations to fuse arithmetic operators. Being from a compiler background, Nate asked whether we could view these as more traditional compiler transformations. In this blog, we’re going to show how we answered this question for a specific example, namely discovering carry-save addition (a hardware optimisation) via loop fusion (a ...

Darkbloom – Private inference on idle Macs

453 twapi 2026-04-16 04:26:21 UTC darkbloom.dev Llama Share 220

Darkbloom is a decentralized AI inference network connecting idle Apple Silicon machines to demand, reducing costs by up to 70% and giving operators 95% of revenue. It ensures end-to-end encryption and operator anonymity, eliminating data extraction risks.

Artifacts: Versioned storage that speaks Git

102 jgrahamc 2026-04-16 13:26:21 UTC blog.cloudflare.com Llama Share 6

Cloudflare introduces Artifacts, a distributed, versioned filesystem built for agents, serving applications with Git-like functionality. Artifacts aims to address the scale and efficiency needs of modern development, with a private beta available for paid Workers plan developers.

Show HN: MacMind – A transformer neural network in HyperCard on a 1989 Macintosh

91 hammer32 2026-04-16 14:26:21 UTC github.com/seanfdz Llama Share 27

MacMind is a 1,216-parameter transformer that learns the bit-reversal permutation from random examples using HyperTalk, a scripting language. It demonstrates that backpropagation and attention are math, not magic, and can be run on a 68000 processor from 1987.

IPv6 traffic crosses the 50% mark

726 Aaronmacaron 2026-04-16 06:26:21 UTC google.com Llama Share 517

We measure IPv6 connectivity among Google users, showing deployment and reliability issues in various regions. Regions with high IPv6 deployment have fewer issues, while those with low deployment experience significant problems.

We gave an AI a 3 year retail lease and asked it to make a profit

160 lukaspetersson 2026-04-16 15:26:21 UTC andonlabs.com Llama Share 233

Andon Labs deployed an AI named Luna to run a retail store in San Francisco, Andon Market, with human employees. Luna made all decisions, including hiring, but hid her AI status from some applicants.

Show HN: CodeBurn – Analyze Claude Code token usage by task

48 agentseal 2026-04-16 16:26:21 UTC github.com/agentseal Llama Share 13

CodeBurn is an interactive dashboard for AI coding tools, supporting multiple providers and models, with features like session data tracking, one-shot success rate, and customizable currency. It provides a TUI dashboard, CSV/JSON export, and macOS menu bar widget.

The paper computer

256 jsomers 2026-04-16 02:26:21 UTC jsomers.net Llama Share 81

The user wants to combine physical and virtual worlds to reduce screen time and distractions, allowing for flexibility and convenience. They envision a future where technology helps people work and interact without constant screen interaction.

Six Characters

63 Airplanepasta 2026-04-16 14:26:21 UTC ajitem.com Llama Share 9

Air India's Passenger Name Record (PNR) is a structured document containing a unique locator and five mandatory elements, used to identify a booking. The PNR is associated with an e-ticket number, which is the primary key and remains stable even when the booking is re-accommodated.

FSF trying to contact Google about spammer sending 10k+ mails from Gmail account

345 pabs3 2026-04-16 04:26:21 UTC daedal.io Llama Share 197

You're looking for a direct contact at Google's Gmail team to report a spammer issue. You've tried the abuse form but want to reach a human employee for a solution.

Cloudflare Email Service

348 jilles 2026-04-16 13:26:21 UTC blog.cloudflare.com Llama Share 154

Cloudflare Email Service is now in public beta, allowing applications and agents to send emails globally with low latency. It unlocks a full email client and agent pipeline for developers.

Japan implements language proficiency requirements for certain visa applicants

84 mikhael 2026-04-16 17:26:21 UTC japantimes.co.jp Llama Share 44

Japan now requires language proficiency for certain foreign nationals applying for the Engineer/Specialist visa. Applicants must show proof of language ability, such as a JLPT N2 certificate or a BJT score of 400 or higher.

Codex Hacked a Samsung TV

175 campuscodi 2026-04-16 11:26:21 UTC blog.calif.io Llama Share 104

Researchers used AI to hack a Samsung TV, partnering with OpenAI, and successfully escalated privileges from a browser shell to root. The AI, Codex, used a physmap primitive to access physical memory and overwrite kernel credentials.

AI cybersecurity is not proof of work

169 surprisetalk 2026-04-16 11:26:21 UTC antirez.com Llama Share 74

Proof of work analogy fails for AI security as bugs are not found through brute force but rather through model intelligence and understanding. Better models and faster access to them will win in AI security, not more computational power.

Modern Microprocessors – A 90-Minute Guide

163 Flex247A 2026-04-16 09:26:21 UTC lighterra.com Llama Share 19

Processors execute instructions in a pipeline, with multiple stages for fetching, decoding, executing, and writing back results, allowing for a four-fold speedup without changing clock speed. Modern processors use superscalar execution, where multiple instructions are executed in parallel, and branch prediction to minimize mispredictions, but still suffer from diminishing returns due to deep ...

European civil servants are being forced off WhatsApp

16 aa_is_op 2026-04-16 19:41:21 UTC politico.eu Full Share 4

New AI tech could pose major cybersecurity risks but many European regulators have limited oversight.

ChatGPT for Excel

311 armcat 2026-04-15 21:26:21 UTC chatgpt.com Full Share 190

Ask ChatGPT to build full spreadsheets, get insights across tabs and formulas, and update workbooks in real time so projects move forward faster. Available globally for ChatGPT Business, Enterprise, Edu, Teachers, and K-12 users, and for ChatGPT Pro and Plus users outside the EU.

PHP 8.6 Closure Optimizations

93 moebrowne 2026-04-16 10:26:21 UTC wiki.php.net Llama Share 21

Two new optimizations for closures are proposed to improve performance by turning non-static closures into static ones and caching stateless closures. This may cause some backward compatibility breaks.

RamAIn (YC W26) Is Hiring

0 2026-04-16 07:26:21 UTC ycombinator.com Llama Share

RamAIn is a YC W26 company that builds AI agents for enterprise work, eliminating repetitive workflows. They're hiring a generalist for go-to-market, sales, and operations to work directly with the founders.