AI dev tools

29 October

DDoSecrets publishes Icepick dataset of 41,000 DHS staff records leaked by hacktivists

Details concerning the names, offices, contact information, and assignments of over 41,000 U.S. Department of Homeland Security (DHS) personnel have surfaced following a data leak.

29 October

Shadowy AI Relay disables OpenAI Codex, Claude Code sandboxes for China

Popular AI coding tools are being made available in mainland China and other territories where OpenAI and Anthropic offerings lack official distribution, though this access comes with crucial…

29 October

Anthropic Claude Code team tests model reliance, claims no mental model needed for codebase

The process for reviewing code within the Claude Code team is now documented, revealing that practitioners operate without maintaining a full mental picture of the entire codebase, a claim Anthropic…

29 October

next.js 16.5.0-canary.31 released

29 October

HPE Alletra MP X10000 offload delivers 20x faster time to first token in AI testing

The Alletra Storage MP X10000 cache offload, as demonstrated by Kamiwaza, achieves up to a 20× faster time to first token, suggesting that remembering information is often more beneficial than…

29 October

Datadog Scout AI model investigates production problems in Bits Investigation

Scout, a specialized artificial intelligence model, is now available from Datadog, designed specifically to aid in the investigation of live production issues.

29 October

AI Token Usage Metric Risks Bloated Code, Misleading Performance in Developer Evaluation

The tendency for developers to inflate their output by using AI excessively mirrors the pitfalls of outdated performance measurements like lines of code.

29 October

Cloudflare releases Web Search API via AI Gateway for AI applications

Developers can now equip their AI agents with the ability to fetch up-to-date web content, such as recent technical documentation, API alterations, and current news, by leveraging Cloudflare's new…

29 October

Why More AI-Generated Tests Don't Mean Better Testing

Every demo from an AI testing tool vendor ends with a promise to cut test design effort from hours to minutes.

29 October

Pwn2Own Berlin 2026: Handmade, logic-only Microsoft Edge Full Chain revealed

A fully custom-built, logic-only Microsoft Edge Full Chain was developed for Pwn2Own Berlin 2026.

29 October

AIOps advances DevOps decision-making, reducing alert fatigue and predicting failures

Many engineers now rely on AI in at least one professional duty, such as coding, debugging, or data analysis, firmly establishing the technology as a central pillar of contemporary DevOps rather…

29 October

Qualys CISO discusses handling AI-discovered vulnerabilities, tracking exposure window

Nathan Smolenski, the CISO at Qualys, discussed how security departments can manage the increasing quantity of vulnerabilities unearthed by artificial intelligence.

29 October

Gemini 4 Carbon, Google, debuts stronger model, undercutting Claude Opus 5.5 price

The engineering team at Google DeepMind has unveiled Gemini 4 Carbon, a model that boasts superior performance while maintaining the identical initial pricing structure as its predecessor, Gemini 4…

29 October

Fujifilm boosts LTO production capacity 5x to meet AI data demands

The need to securely and efficiently store and utilize the massive quantities of data generated through AI training and inference, alongside the digitalization of business processes, is becoming…

29 October

Alphabet Q3 earnings show AI Mode searches earn less than old search ads

The company is currently forced to rent computing power from outside providers because it has invested $51 billion in its own infrastructure, illustrating a current hurdle for its AI initiatives.

29 October

Qwenputer: A functioning, deterministic RISC-V CPU inside Qwen3.5

I compiled a RISC-V CPU into Qwen 3.5 4B's weights, using its Gated DeltaNet layers as RAM.

28 October

EMACS turns 50

How the MIT AI Lab’s text editor gave birth to the free software movement.

28 October

Ted Sanders explains OpenAI model versioning, size impacts performance, cost, and recall

The way these large language models perform is remarkably complex, meaning there is no fixed hierarchy where the model excelling at mathematics will automatically be the one best suited for humor.

28 October

Google, NASA partner to build robot medic for deep-space astronaut assistance

The robot is designed to function as an extra set of hands for a human doctor, rather than a complete digital medical officer.

28 October

GitHub releases Copilot tools to manage parallel AI coding agents, adds Copilot Fusion

Enterprise customers will no longer face an additional licensing fee for utilizing GitHub's agentic AI tools beyond the expenses associated with running the underlying AI models.

28 October

Anthropic engineer Felix Rieseberg stops using coding interviews, citing AI capabilities

The belief that the coding interview is obsolete is being put forward by an engineer at Anthropic, who argues that asking humans to complete coding tests is redundant when AI tools can handle them.

28 October

Next.js 16.4 adds next-bundle-optimizer agent skill to reduce JavaScript bundles

Agents are now equipped with the capability to streamline JavaScript delivery in Next.js applications using the next-bundle-optimizer agent skill.

28 October

Vercel CI is now in beta

You can now run CI for your Turborepo with one line of configuration.

28 October

European digital rights nonprofit noyb sues Scale AI for ‘invasive’ worker surveillance

The Austria-based digital privacy advocacy group noyb on Tuesday filed five complaints against a major artificial intelligence company alleging that its practices violate European law by tracking freelance workers’ web activity as they generate AI model training data.

28 October

Box Mount now runs in Vercel Sandbox

Box Mount lets agents running in Vercel Sandbox read and write Box files through standard filesystem operations.