XL Today - Spreadsheets. Data. Now.

BREAKING

Evaluation reveals AI confidence peaks on errors | New AI model uncovers major Cursor flaw | DeepSeek Harness Debuts Open-Source Rival to Claude Code | AI Agent Identity Solved Containment Proves Difficult | AI video made in under 7 seconds | OpenAI unveils GPT‑5.6 Cyber with 95% success | Stanford AI agents design drug built by Merck | Four AI Agents Outperform Claude Opus 4.8

Spreadsheet Audits

Evaluation reveals AI confidence peaks on errors

Danielle Hayes · 49 minutes ago

Top Stories See All »

New AI model uncovers major Cursor flaw

New AI model uncovers major Cursor flaw

DeepSeek Harness Debuts Open-Source Rival to Claude Code

Automation Signals

DeepSeek Harness Debuts Open-Source Rival to Claude Code

AI Agent Identity Solved Containment Proves Difficult

Formula Workshop

AI Agent Identity Solved Containment Proves Difficult

AI video made in under 7 seconds

Spreadsheet Audits

AI video made in under 7 seconds

Latest News

Automation Signals

OpenAI unveils GPT‑5.6 Cyber with 95% success

Heather Reed · 5 days ago

Productivity Hacks

Stanford AI agents design drug built by Merck

Amber Sull · 6 days ago

Automation Signals

Four AI Agents Outperform Claude Opus 4.8

Heather Reed · 1 week ago

Data Workflow

Why Qwen and Opus 5 scores miss the cost mark

Amber Sull · 1 week ago

Automation Signals

Liquid AI launches LFM2.5 model for Raspberry Pi

Heather Reed · 1 week ago

Data Workflow

AI reveals network architecture flaws

Danielle Hayes · 1 week ago

Spreadsheet Audits

Shai-Hulud worm passes security check legitimately

Heather Reed · 1 week ago

Productivity Hacks

Commerce AI’s Hidden Metric Flaw Sparks Industry Concern

Amber Sull · 2 weeks ago

Automation Signals See All »

DeepSeek Harness Debuts Open-Source Rival to Claude Code

OpenAI unveils GPT‑5.6 Cyber with 95% success

Four AI Agents Outperform Claude Opus 4.8

Spreadsheet Audits See All »

Evaluation reveals AI confidence peaks on errors

AI video made in under 7 seconds

Shai-Hulud worm passes security check legitimately

Productivity Hacks See All »

Stanford AI agents design drug built by Merck

Commerce AI’s Hidden Metric Flaw Sparks Industry Concern

DataFlow Harness Bridges AI Pipeline Accuracy Gap

Formula Workshop See All »

AI Agent Identity Solved Containment Proves Difficult

GraphRAG outperforms vector method in some cases

Mastercard updates fraud rules for AI assistants

Data Workflow See All »

New AI model uncovers major Cursor flaw

Why Qwen and Opus 5 scores miss the cost mark

AI reveals network architecture flaws