groundy

Groundy — independent coverage of developer tools, infrastructure, and platforms

policy

AI Agent Alignment Tests Are One-Shot. A New Benchmark Catches Multi-Step Failures

MoralityGym proves AI agents pass one-shot alignment checks but drift toward violations across multi-step trajectories, a failure mode red-team prompt batteries cannot detect.

6min · 17:09 · may 23
oss

Files.md Bets on Plain Markdown Folders as the Obsidian Exit Ramp

Files.md's HN traction exposed that Obsidian's core is closed-source, reframing its plugin ecosystem as a migration tax for developers evaluating plain-markdown alternatives.

6min · 16:34 · may 23
industry

Green Card Rule Change Forces Tech Workers to Leave the US to Apply

A May 22 USCIS memo eliminates standard in-country green card processing, forcing temporary visa holders into consular processing abroad with no guaranteed return.

culture

Microsoft's Own Numbers: AI Agents Cost More Per Task Than the Human Employees They Replace

Microsoft's internal data shows agentic AI workflows consume 1000x more tokens than chat, with costs varying 30x between identical runs and zero correlation to output quality.

policy

Microsoft's Own Numbers Now Show AI Agents Cost More Than the Humans They Replaced

Microsoft's internal data shows token-burning AI agents now exceed the all-in cost of human labor, giving procurement teams vendor-supplied evidence to challenge 2027 renewal.

industry

OpenAI Hires Slack's Denise Dresser as CRO, Conceding Enterprise Growth Needs a Sales Org

OpenAI hiring Salesforce veteran Denise Dresser as CRO signals a shift from product-led growth to field sales, driven by IPO pressure and Anthropic's enterprise momentum.

security

OpenAI Ships Lockdown Mode and Elevated Risk Labels for ChatGPT Sessions

OpenAI's Lockdown Mode kills ChatGPT network exfiltration paths at the infrastructure layer, conceding that model-level filtering cannot stop prompt injection at scale.

culture

Trump Ends Domestic Green Card Filing: Applicants Must Now Leave the US to Apply

A May 22 USCIS memo closes the domestic adjustment-of-status path, requiring H-1B, L-1, and F-1 visa holders to leave the US and refile through consular processing abroad.

most popular last 6 days · by traffic
  1. 01 devtools GitHub Copilot vs Cursor vs Claude Code: The 2026 AI Coding Showdown
  2. 02 devtools Claude Code Plugins: Anthropic's Official Plugin Ecosystem Explained
  3. 03 models Chinese AI Models Compared: DeepSeek, Qwen, Kimi, Doubao, and Ernie
  4. 04 infra MLX vs llama.cpp on Apple Silicon: Which Runtime to Use for Local LLM Inference
  5. 05 models AI Code Generation Benchmarks 2026: Which Model Actually Writes Better Code?
  6. 06 infra Prefill-Decode Disaggregation: The Architecture Shift Redefining LLM Serving at Scale
  7. 07 policy Atlassian Turned On AI Training Data Collection by Default — Here's What to Disable
  8. 08 industry Cursor's Meteoric Rise: Inside the AI Editor Hitting $300M ARR
  9. 09 devtools Claude Code in GitHub Actions: A Complete Guide to Automated PR Fixes
  10. 10 devtools GitHub Copilot's Opus 4.7 Multiplier: 7.5x to 15x to 27x in 60 Days
agents more →
devtools more →
security more →
industry more →
feed all 303 articles · reverse chronological
  1. may 23 policy AI Agent Alignment Tests Are One-Shot. A New Benchmark Catches Multi-Step Failures
  2. may 23 oss Files.md Bets on Plain Markdown Folders as the Obsidian Exit Ramp
  3. may 23 industry Green Card Rule Change Forces Tech Workers to Leave the US to Apply
  4. may 23 culture Microsoft's Own Numbers: AI Agents Cost More Per Task Than the Human Employees They Replace
  5. may 23 policy Microsoft's Own Numbers Now Show AI Agents Cost More Than the Humans They Replaced
  6. may 23 industry OpenAI Hires Slack's Denise Dresser as CRO, Conceding Enterprise Growth Needs a Sales Org
  7. may 23 security OpenAI Ships Lockdown Mode and Elevated Risk Labels for ChatGPT Sessions
  8. may 23 culture Trump Ends Domestic Green Card Filing: Applicants Must Now Leave the US to Apply
  9. may 23 culture US Researchers Hit With New Federal Limits on Publishing With Foreign Collaborators
  10. may 23 infra What Cloudflare's Q1 2026 Outage Data Says About Designing for State-Level Shutdowns
  11. may 22 models A Theory of Time-Sensitive Language Generation Says Sparse Hallucination Beats Mode Collapse
  12. may 22 models arXiv 2605.16428 Measures AI Search's Drag on Publisher Traffic Using Paired Google and Reddit Data
  13. may 22 agents Beyond Text-to-SQL: New Agentic Architecture Routes Enterprise Analytics Through Governed APIs
  14. may 22 policy CISA's Own Data Leak Has Lawmakers Demanding Answers About the Voluntary Threat-Sharing Pact
  15. may 22 devtools Cursor's In-House Model Changes the Vendor Calculus for AI Coding Teams
  16. may 22 devtools Deno 2.8 Lands as Bun Gets Deprecated by yt-dlp: The JavaScript Runtime Field Is Reshuffling
  17. may 22 culture Employer-Side Law Firms Create a Structural Asymmetry in US Organizing Drives
  18. may 22 devtools Google Sunsets Gemini CLI on June 18: Forced Migration to Antigravity CLI Breaks Existing Automation
  19. may 22 agents GraphFlow Lifts LLM-Agent Workflows Into Schedulable Graphs to Optimize Serving
  20. may 22 agents Learning to Configure Agentic AI Systems Exposes a Gap in CrewAI and AutoGen Template Libraries
  21. may 22 devtools Malicious VSCode Extension Hit 3,800 Repos: What GitHub's Marketplace Trust Model Actually Verifies
  22. may 22 security AI Jailbreaks Are Now a Reasoning Problem, Not a Prompt Problem
  23. may 22 industry Microsoft and Uber's AI Agent Bills Expose a Per-Token Pricing Problem
  24. may 22 agents Microsoft's 2026 Cost Math Forces CrewAI and LangGraph Users to Audit Token Spend Per Agent
  25. may 22 policy NIH Demands Advance Clearance for Foreign Co-Authors Without a Published Rule
  26. may 22 oss Nx Console 18.95.0 Compromise Hides a Multi-Stage Credential Stealer in an Orphan Commit
  27. may 22 security OpenAI's New Agent Defense Post Concedes Prompt Injection Is Architectural, Not Patchable
  28. may 22 industry OpenAI's S-1 Will Force the First Public Audit of LLM Inference Margins
  29. may 22 industry OpenAI's S-1 Will Have to Define AGI for SEC Reviewers, Not Just Investors
  30. may 22 agents PBT-Bench Asks Whether AI Coding Agents Can Actually Write Property-Based Tests
  31. may 22 infra Railway's May 19 GCP Suspension Exposes the Single-Account Risk Underneath Every Reseller PaaS
  32. may 22 security Jailbreak Defense Now Lives in Model Weights, Not in Prompt Filters
  33. may 22 agents AI Agents That Learn New Skills Without a Human Curator
  34. may 22 agents SpecBench Catches Long-Horizon Coding Agents Gaming Reward Signals
  35. may 22 agents SpecBench Exposes Reward Hacking in Long-Horizon Coding Agents
  36. may 22 security Vercel Blocks Deploys With Vulnerable next-mdx-remote by Default: Platform Mitigation Outpaces the CVE Cycle
  37. may 22 security Vercel's Next.js Middleware Bypass Postmortem: What the Fix Reveals About Edge Runtime Auth
  38. may 22 infra vLLM 0.21 Makes Prefill-Decode Disaggregation Actually Practical
  39. may 22 security When Stronger Backdoor Triggers Backfire: An arXiv Theory Paper Inverts a Core Defense Assumption
  40. may 18 agents A New Trust Schema Exposes Why Agent Skill Registries Fail Enterprise Audit Requirements
  41. may 18 industry Anthropic Passes OpenAI in US Business Adoption, But Per-Token Billing Shifts Cost Risk to Buyers
  42. may 18 industry Anthropic Ships 10 Finance Agents With Moody's 600M-Company Credit Data and Expanded Microsoft 365 Integration
  43. may 18 industry Bret Taylor's Sierra Raises $950M at $15B, Claims 40% of Fortune 50 Use Its Agents
  44. may 18 infra DMax Hits 1,338 Tokens/Sec on 2x H200: Parallel Decoding Pushes dLLM Serving Past the Autoregressive Bar
  45. may 18 infra KV Cache Offloading Breaks on Text2JSON: Why Llama 3 and Qwen 3 Lose Accuracy on Context-Intensive Prompts
  46. may 18 policy Maryland Enacts First US Ban on Algorithmic Grocery Pricing, Effective Immediately
  47. may 18 models The Last Word Often Wins: A Format Confound Inflates Chain-of-Thought Corruption Robustness Scores
  48. may 18 agents Trojan Hippo Plants Dormant Payloads in Agent Memory, Hits 85-100% Exfiltration on Frontier Models
  49. may 17 culture AB 566 Forces Chrome and Safari to Ship Opt-Out Signals by 2027 — Then Shields Them from Google's 86% GPC Failure
  50. may 17 industry AI Was Cited in 26% of Challenger's April Layoffs. UBS Notes the Series Captures 5% of US Job Flow
load older →