Build a private AI content scoring app with a rubric and human review
Build a private team content scoring app with versioned drafts and rubrics, saved model estimates, exact historical inputs, and separate human review.
The BIG CHANGE archive
News, analysis and practical guides.
Explore our reporting on how AI and robotics change work, business and the world.
The latest change
849 stories
Finish an existing Replit Agent app and check its first release
Audit an unfinished Replit Agent app against its original requirements, finish one user journey, then verify the whole project's first restricted release.
Add sign-in and protected data to an existing React app
Use a coding agent to add Supabase email authentication to an existing React/Vite app, then check real callbacks, enrollment and per-user data access before release.
Build a private dashboard from real data with Replit Agent
Define a real event export, build its database, import, API and UI with Replit Agent, then reconcile metrics and check an Invite only release.
Build a source-backed LLM wiki and check its first update
Build linked Markdown pages from two permitted sources, check an AI answer and a knowledge gap against originals, then audit and repair the wiki.
Install Claude Code on Mac or Windows and verify your first project read
Install the native Claude Code CLI on Mac or Windows, verify its setup and account, then check one authorized read in your own project.
Build a local AI product copy review app and check every claim
Build a single-user Python app that reviews product copy against approved facts, shows quoted issues, and saves the human-edited version in local SQLite.
Build a feature feedback board with Replit Agent and check its release
Give Replit Agent a complete feature board brief, then inspect identity, persistent requests, unique votes, moderation and a checked production release.
Move a retired Claude Sonnet 4 API application to Sonnet 5.5 and check the result
Migrate one owned, nonstreaming Claude Messages API application from retired Sonnet 4 to Sonnet 5.5, then check its real task, tools, effects, usage and release decision.
Make a checked Gmail follow-up draft from Notion and HubSpot
Connect Notion, HubSpot and Gmail to Claude Code, reconcile an approved brief with exact CRM and email records, and verify one human-reviewed Gmail follow-up draft.
Verify one OpenAI Agents SDK tool boundary before release
Test one existing Python Agents SDK FunctionTool against approved and denied arguments, then review an authorized staging effect, trace and usage for a bounded accept-or-hold decision.
Run two Claude Code Desktop tickets in separate projects
Set up two local Claude Code Desktop worktree sessions for unrelated Git repositories, then verify each ticket's permissions, diff and tests independently.
Correct Claude Code project memory and check a fresh session
Inspect and correct one owned Claude Code project memory note and its index, then check a fresh local session against the project's authoritative source.
Salesforce agrees to acquire Braze for $5.66 billion
The proposed acquisition would bring Braze's real-time customer engagement and AI decisioning into Salesforce's platform. The $5.66 billion agreement still needs Braze shareholder and regulatory approval.
Chrome adds a suspicious-site warning for Enhanced Protection users
Chrome now separates lower-confidence suspicious sites from known dangerous ones in Enhanced Protection. Google's Gemini scam analysis and ClickFix clipboard checks add detection, with server processing and privacy conditions users should understand.
Set and check Vercel spend alerts for an AI-built app
Set a Vercel team spend amount, choose your alert thresholds, and record the AI app project's usage without changing existing pause or webhook actions.
Check README npm scripts with a Claude Code cloud scheduled task
Set up one Claude Code cloud scheduled task to compare README npm run commands with package.json scripts, then verify manual and scheduled reports at their checked commits.
Turn owned HTML, PDF and Word files into checked Markdown context
Convert permitted local HTML, PDF and Word files to Markdown, check material content against originals, assemble a source-mapped AI context packet and count comparable text tokens only when a valid baseline exists.