Articles from X
Blogger vanillaCitron shares their initial impressions after spending 30 minutes with the iPhone Duo, along with a link to the full write-up.
Aaron Levie argues that AI-driven productivity divergence is at an all-time high: agentic models, better harnesses and data have pushed frontier gains from 30-50% to potentially 5-10X in two years, widening the gap between frontier users and everyone else as RSI accelerates.
A tweet speculates that OpenAI's Tibo is hyping GPT-6.2 'Bel' and suggests a major AI release could drop today.
A look at Labelme, an open-source Python/Qt image annotation tool for drawing polygons, boxes, circles, and points, featuring AI-assisted annotation via SAM, EfficientSAM, YOLO-World, and SAM3, plus VOC/CCO-format exports and video annotation support.
Binance reports that hourly deposits have exceeded withdrawals as of 03:00 UTC, indicating improving user activity and stabilizing fund flows on the exchange.
MIT Press's 634-page book "Algorithms for Optimization" is available as a free CC-BY-NC-ND PDF, with the second edition (featuring Julia implementations and new chapters on DCP, surrogate models, and multidisciplinary optimization) releasing January 5, 2027.
The author agrees with poteto, criticizing adversarial code review between sub-agents as a waste of tokens, pointing out its three fatal flaws — meaningless nitpicking, over-engineering, and mismatched compute — and noting that top teams are shifting toward real testing, formal verification, and actual red-teaming to find issues.
RL can assist robot data collection during demonstrations through a technique called Data Inoculation, improving data quality for both behavior cloning and RL, as shown in a new paper on stacking cubes.
China Medical University Hospital, Taiwan's second-largest healthcare provider, uses VMware Cloud Foundation and vDefend to deploy compliant IT in under half a day while protecting clinical workloads with Zero Trust security.
A user shares how Grok Bot autonomously shopped for auto insurance and called Progressive to inquire about glass coverage while they were driving, highlighting Grok's agentic capabilities.
A UIUC student is seeking compute sponsors for Loxeon, an Apache-2.0-to-be runtime where AI agent teams self-organize via formally specified primitives (identity, capability delegation, role rebinding, tree-based management, and epoch/ticket message validation), written in Rust and checked against a formal spec. The goal is to fairly benchmark whether multi-agent teams outperform single agents on real coding tasks (~$200 small / $1,500 full), with code, logs and results to be published regardless of outcome.
GitHub thanks attendees for joining GitHub Universe 2026 and looks forward to what developers will build next.
A user publicly accuses Railway employee Jake Runzer of attempting to breach their infrastructure, sparking a security-related allegation against the hosting platform.
An introduction to the open-source Agent Skill "learn-from-materials," which converts hundreds of pages of material — PDF, EPUB, Word, PPT, web pages, and more — into an interactive learning website and knowledge base, with citation tracing precise down to page numbers and chapter sections. Compatible with environments like Claude Code, Codex, and Copilot CLI.
推荐一个为 OpenWrt 25.12/24.10 从头构建的 LuCI 主题与实时网络仪表盘「Vantage」，包含主题 luci-theme-vantage 和应用 luci-app-vantage，提供健康检查、网络路径可视化、实时吞吐量与客户端管理等功能。
Karpathy used Claude to analyze 43 million Hacker News comments from the past 20 years, identify 6,454 repeatedly recommended videos, and rebuild HackerTube into a better YouTube recommendation tool in about 2 hours.
A Google and University of Hong Kong paper proposes VeriCUA, which verifies computer-use agent tasks by checking saved files and app state instead of trusting the agent's own completion report, cutting false positives to 5.4% versus 24.8% for a VLM judge.
OpenAI is more than quadrupling bug bounty reward ranges for medium, high, and critical vulnerabilities, adding quality multipliers, a tiered loyalty program, priority campaign bonuses, and retroactive payments for recent awards.
OpenAI announced its next internal security sprint, partnering with Cathedral Systems to use its models to find and fix vulnerabilities, while quadrupling bug bounty rewards with critical findings eligible for up to $172,000.
A collection of 6 GitHub quant trading projects that connect the learning → strategy → backtesting → live trading pipeline: awesome-quant, machine-learning-for-trading, quant-trading, backtrader, zipline, and ccxt.