Research report 10 / Reader behavior

AI Attention vs. Substance Index

Reader attention plotted against editorial importance, source depth, completion, saves, and shares.

Archived snapshotv140Oct 4, 2026
Measured stories
500

With first-party analytics

Total opens
47.5K

Across measured stories

Median completion
22.4%

Completions divided by opens

High-intent actions
95

Saves and shares

Interactive figureAttention vs. substance
CSV JSON
Data status500 verified observations

Each dot is one verified observation. Horizontal position is editorial importance score; vertical position is completion rate (%).

Verified observationHover or focus any mark for exact valuesLast updated Oct 4, 2026

Version ledger

Frozen public editions

Each edition preserves the records, method, sources, and downloads available at publication time.

  1. v141 / latestOct 5, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  2. v140Oct 4, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  3. v139Oct 3, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  4. v138Oct 2, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  5. v137Oct 1, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  6. v136Sep 30, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  7. v135Sep 29, 2026500 records / 0 sources

    updated Total opens, High-intent actions. The frozen snapshot contains 500 records.

  8. v134Sep 28, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  9. v133Sep 27, 2026500 records / 0 sources

    updated Total opens, High-intent actions. The frozen snapshot contains 500 records.

  10. v132Sep 26, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  11. v131Sep 25, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  12. v130Sep 24, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  13. v129Sep 23, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  14. v128Sep 22, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  15. v127Sep 21, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  16. v126Sep 20, 2026500 records / 0 sources

    updated Total opens, High-intent actions. The frozen snapshot contains 500 records.

  17. v125Sep 19, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  18. v124Sep 18, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  19. v123Sep 17, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  20. v122Sep 16, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  21. v121Sep 15, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  22. v120Sep 14, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  23. v119Sep 13, 2026500 records / 0 sources

    updated Total opens, Median completion. The frozen snapshot contains 500 records.

  24. v118Sep 12, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  25. v117Sep 11, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  26. v116Sep 10, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  27. v115Sep 9, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

  28. v114Sep 8, 2026500 records / 0 sources

    updated Total opens. The frozen snapshot contains 500 records.

  29. v113Sep 7, 2026500 records / 0 sources

    updated Total opens, High-intent actions. The frozen snapshot contains 500 records.

  30. v112Sep 6, 2026500 records / 0 sources

    updated Total opens, Median completion, High-intent actions. The frozen snapshot contains 500 records.

Coverage note

Reader analytics are aggregated and privacy-safe. The audience and editorial selection create known sampling bias, disclosed here rather than normalized away.

Dataset ID
spd:attention-substance
Stable URL
/research/attention-substance
Version
v140
Coverage
Live collection
Records
500
Fields
8
Updated

Read the data

The records behind the figure

CSV JSON
AI Attention vs. Substance Index data records
StorySignalImportanceImpressionsOpensCompletion rateSavesShares
Win 1 year free Pro subscription to Superpower ChatGPT ($200 value) 🤑—50161873.300
San Francisco Sends Meta Cease-and-Desist Over AI Child Abuse Adsgovernment-action8230952311.501
Developers Turn a New Fruit Fly Brain Map Into Viral Simulationsresearch634150522.804
MIT’s η-Learning Generates 100-Year Storm Maps Without Training on Extreme Eventsresearch704642219.204
OpenAI Codex Client Shows Luna Reserve Fallback After Advanced-Model Limits—604336234.800
OpenAI Halts New $200 Pro Sign-Ups as Astra Demand Strains Capacityplatform-shift7213036030.301
OpenAI Pauses Frontier Training After Sandbox Breach, Warns of Persistent AI Cyberattackssecurity-risk901373316.900
OpenAI Slows Reinforcement Learning for Two Weeks After Agents Breached Hugging Facesecurity-risk90843304.500
Anthropic’s Parka Design Would Turn Meetings Into Claude Agent Tasksresearch828229525.100
OpenAI Launches Astra for Law as a Platform for Legal AI Builderslaunch7611129528.502
Flock Safety Employee Buyouts: Deadline and Severance Details—554229340.300
OpenAI DevDay is today: when and how to watch—75127928420.400
AT&T Expects More Job Cuts but Rejects Reported 85,000-Worker Target—582227130.304
Bracket22: Brian Kelly's AI Trading Firm Keeps a Human in Chargeenterprise-adoption628726940.500
Everything OpenAI announced at the DevDay 2026 keynoteplatform-shift94191726511.300
Baseten Brings Open Models to OpenAI Enterprise Customers Under Existing Commitmentspartnership606823724.900
Meta’s Muse tops ChatGPT in early iOS downloads——552296.602
OpenAI says new GPT-6 models cost half as much——672277.911
MIT Weighs Education Overhaul as AI Handles Almost Any Written Undergraduate Assignment—706221931.101
Harvard puts AI professors in a $699 bootcamp—50982187.300
Anthropic weighs a model as Astra gains ground——552165.600
EU Seeks AI Act Answers From 30-Plus Companies, With Fines for Misleading Repliesgovernment-action8011621214.600
ChatGPT Gets Into Apple Messages—5017421115.601
OpenAI’s 10-Day WebMCP Challenge Pushes Websites to Expose Tools for AI Agentslaunch8010620611.200
Altman Warns AI Power May Concentrate as He Revises Disruption Timeline—707620410.800
OpenAI Releases GPT-6 Sol and Luna With API Prices Half Their Predecessors’launch8830520340.400
OpenAI Adds Role-Based Academy Courses With Assessments for AI Skillslaunch584720213.901
OpenAI releases GPT-6 Astra with new safeguards——6719911.600
MIT Study Finds Chatbot Help Can Erode Fake-News Detection Without AIresearch804519710.200
Cloudflare Open-Sources Cloudflare OS, Putting AI-Built Work Apps Behind Permission Gatesopen-release801491968.701
Plaintiff Says Anthropic’s $200 Claude Plan Delivers 6x Weekly Sonnet Use, Not 20xlawsuit805719519.500
AI Travel Referrals Rose 119% as the Booking Gap Nearly Disappearedresearch709819514.900
OpenAI Adds $100-Annual ChatGPT Business Seats With 5x Usage for Heavy Userspricing-change80701921300
OpenAI Reportedly Tests Pay-on-Completion AI Deals With Some Large Customerspricing-change7011118710.700
OpenAI Connects ChatGPT Health to Epic, Keeping Clinical Record Access Read-Onlypartnership807218713.401
Higgsfield Puts Its Video Engine Behind Grok Bot—and Keeps the Meter Runningpartnership7010218710.200
OpenAI Releases GPT-6 Astra With Its First Advanced Cyber Safeguardslaunch659718612.920
London Surgeons Remove 11mm Brain Tumour With Live AI Anatomy Mappingresearch708118514.100
Anthropic Commits $100 Million to Train 10,000 Engineers for Claude Deploymentslaunch6466718221.400
Meta’s Muse Beats ChatGPT’s Early Mobile Downloads in Apptopia Estimates—584518138.100
Meta Introduces $1,299 VR Glasses and a Handheld Shortcut to Its AI Agentlaunch635818030.600
Apple Introduces iPhone Duo With an AI-Assisted Hingelaunch668217930.700
How to (almost) protect the instructions of your CustomGPT—5031794.500
Harvard Uses 7 AI Faculty Avatars in a $699 Founder Bootcamp, While Keeping Live Advice Humanlaunch707717819.700
GPT-6 Astra Hits 99.9% on ARC-AGI-3, but Scores 62.7% in a Shared Testlaunch851481762100
Meta Ran Ads for a Tool Promoting Sexualized Political Deepfakessecurity-risk809717611.400
World Labs Launches Atlas for 1440p Camera-Controlled Video, 3D Worlds and Robot Viewslaunch809917517.100
Apple Refreshes Mac mini and Mac Studio for Linked Local AI, From $899launch804017310.400
Oxford Researchers Catch AI Agents Using Coded Blackjack Messages to Colluderesearch669117344.503
OpenAI slows reinforcement learning after agent breach—50741728.700
OpenAI Pauses Frontier Training as AI Controls Move to the Forefront—507917111.700
Fable 5.1 leads, but its top setting costs more——641698.300
Anthropic Adds a Coordinator for Parallel Claude Code Sessionslaunch7011216915.400
Anthropic Weighs New AI Model as OpenAI’s Astra Gains Ground—6210116842.900
Claude, Codex and Hermes Ran Unowned Package Commands Inside Corporate Networkssecurity-risk9014216816.100
PitchFly: WIRED's Fruit Fly Brain Experiment for Story Ideas—4210416820.801
OpenAI Partners With Local Business Advisors to Bring AI Training to 1,000 Small Businessespartnership576116638.601
Stripe’s OpenRouter Deal Puts AI Routing Beside the Billing Stackacquisition9017816516.400
Ant Puts Its FX Forecasting Model Into Tools Used by Major Banksenterprise-adoption80701657.300
Perplexity’s Agent API Adds Managed Connectors for GitHub Code and PR Workflowslaunch801081644.900
Snap pairs $2,195 glasses with an AI assistant——631624.901
Beijing Humanoid Games Put Practical Work to the Test; Only 3 of 12 Finish Fire Coursebenchmark70461618.100
OpenAI Adds Clothing Try-On and Saved Shopping Picks to ChatGPTlaunch5541316023.800
Claude and Codex ran unowned code at companies—506815710.800
Snap Launches an AI Assistant Alongside Its $2,195 AR Glasseslaunch6515815538.700
Anthropic Sets Up Bay Area Wet Lab for Physical Biology Work—6810115512.900
Steven Strogatz Says AI Math Is Pushing Humans Toward Interpretation—6815915443.501
China’s $226M Underwater Data Center Is Running. The Ocean Still Has to Prove It Can Scalelaunch859615313.700
Google's AI Now Delivers Precise 24-Hour Weather Forecasts ⛅—5001493.400
Apple Releases macOS 27 With a Siri That Can Read Your Screenlaunch7417814826.401
Google Research Introduces AI Systems for Longer, More Consistent Videosresearch6314514837.800
Google wants AI videos to stop changing between scenes——521479.500
Hard Fork Continues as Newton and Roose Launch Machine Godspartnership484314633.600
Kevin Bass's Anthropic-METR Funding Claim: What the Audit Found—422814629.500
OpenAI Lets Codex Route Smaller Tasks From Sol to Lower-Cost Luna Workerslaunch70691453101
Amazon’s Las Vegas Warehouse Reportedly Cuts Up Books for AI Training Data—703614515.900
Nvidia AI Server Prices May Rise More Than 15% as Memory Costs Climbpricing-change701331424.900
MIT Builds PottsMPNN to Model Protein Stability Beyond Native-Sequence Matchingresearch80631428.500
Bill Gates Calls for Human-Reserved Jobs and an International AI Framework—704214115.600
OpenAI’s Astra Is Claimed to Solve Non-Sofic Groups Problem, Raising Stakes for Human Mathematiciansresearch805414020.701
OpenAI Says an Astra Training Model Inserted Its Own Jailbreak Instructionssecurity-risk7612514014.300
Leaked: all the details about Humane’s AI Pin—5001398.600
Four Major AI Services Suffer Rare, Overlapping Outagesviral689413922.300
Hugging Face Sells a $399 Microduck Robot With Its Training Stack on GitHublaunch806213919.400
Altman Targets Internal AGI by End of 2026, as Astra Tests OpenAI’s Definitionresearch858613813.800
Tesla Asks Businesses About Buying and Running Cybercab Fleets—644413716.800
OpenAI May Delay a Reported Hodge Conjecture Result After Math Backlash—53671363900
OpenAI Reportedly Buys Tens of Thousands of Macs for Computer-Use Agent Trainingacquisition707613620.600
Google Pics Puts Nano Banana Image Editing Into Docs and Slideslaunch803413417.200
Tilly Norwood Switches to Cantonese While Explaining Its All-Digital Film Castviral413513416.402
Hacktron Says Claude-Assisted Chain Reached OpenAI’s Internal GitHub Environmentsecurity-risk781801346.700
Stanford Finds a 19% Employment Gap for Young Workers in AI-Exposed Jobsresearch807613315.800
Google Provides 100,000 AI Training Scholarships in More Than 80 Countriespartnership653313321.800
Databricks Buys Row Zero to Bring Governed Spreadsheets Into Genieacquisition6821113132.100
Scale AI and Google Cloud Publish Enterprise Agent Deployment Blueprintpartnership641613129.800
OpenAI Tests Persistent Codex Mode That Keeps Working Until Put to Sleepresearch8012913016.900
Sonos 27 Lets ChatGPT Control Music Across Its Speaker Systemplatform-shift58671299.300
Can GPT-4 teach a robot hand to do pen-spinning tricks better than you do?—5001292.300
TypeSafe Launches Jev for Fast, Structured AI Decisionslaunch671041282501
Apple alleges OpenAI used its circuit schematic—506912813.301
Freelancer.com Sees AI-Cleanup Listings Rise 87% as Creatives Shift to Repair Workresearch70481272200
Brood War Bench Finds AI Agents Can Win Matches but Not Master Strategybenchmark526312516.800
Stanford Publishes Paper2Agent to Turn Research Papers Into Working AI Toolsresearch68621251600
OpenAI lets Codex send tasks to cheaper workers—508812318.700
OpenAI Teases Astra With 1979 Voice-and-Gesture Demo, but Leaves Product Details Out—684812330.100
Anthropic’s Fable 5.1 Takes Benchmark Lead, but Its Top Setting Costs More Per Taskbenchmark808912319.500
Job Seeker Sends ChatGPT to AI Recruiter After Five Interviews Without Follow-Up—502512229.500
OpenAI Lets ChatGPT Work Build Writing Profiles From Connected Appslaunch6812812112.400
OpenAI Says 1,000 AI Agents Produced a Navier-Stokes Solutionresearch8310312120.700
Investigation Finds More Than 1,050 Deaths Within Advertised Range of U.S. Border Towers—776112113.200
Meta Says Muse Was Built From Scratch Despite OpenClaw Inspirationviral643612014.200
Anthropic Releases 2030 AI Economy Explorer With a Stark Split in Who Gainsresearch727312019.200
LangChain Adds Jev Classifier to Agent Workflows for Routing and Tool Checkspartnership6710012035.800
Comscore Finds ChatGPT Losing Prompt Share as AI Discovery Splintersresearch686611911.800
100 Firms Warn AI Cyberattacks Will Grow Within Months, Putting Hospitals and Utilities at Risksecurity-risk801241181100
Google’s WikiSkill Lifts Agent Benchmarks by Giving Models a Memory of Failed Workbenchmark8013111816.100
OpenClaw 2.0 Adds Shared Agent Sessions, With No Tenant Isolationlaunch807411816.100
Character.AI Launches Image Models for Consistent Fan Storieslaunch648611711.100
Google DeepMind’s 100-Agent Math Swarm Spread Fake Proofs in 27 Minutesresearch638411733.300
Meta Restricts Staff Access to Muse, but Stronger Privacy Is Still in Testingresearch9299911723.900
OpenAI’s 10-Day WebMCP Challenge Pushes Websites to Expose Tools for AI Agentslaunch70901171200
Rumor: ChatGPT is getting a memory—5031163.400
Figure Tests Helix 2.5 in 30 Unseen Homes, Reports 56% Zero-Shot Successlaunch6712911618.100
Jensen Huang Rejects Robot Tax as Bill Gates Warns AI Could Erode Jobs and Tax Revenuegovernment-action808111616.400
CNN Reports Chatbot-Assisted False Assessment Nearly Led U.S. to Board Chinese Shipsecurity-risk8524111518.300
OpenAI Publishes a Leaner Prompting Playbook for Codex Agents—5716211519.100
Anthropic Opens 750,000 Claude Conversations to Outside Study Without Showing the Chatsresearch8078113800
OpenAI Tests Cross-Session Safety Checks Without Prompt Accesslaunch801171131501
OpenAI in talk to launch the "iPhone of Artificial Intelligence"—5001134.400
Anthropic's Claude Helps Find an Elliptic Curve of Rank at Least 31research665611151.400
Google and Gates Put $100 Million Toward AI Tools for 200 Million Farmerspartnership7215811117.101
AI ECG Tool Flags Heart Risk in Two Seconds, but Still Needs the Scan That Diagnoses Itresearch708011115.300
Meta Halted an AI Layoff Wave After Agents Fell Short of the Planned Team Cuts—804711014.501
Anthropic Adds Claude Code Workflows to Build AI Tests and Tune Apps Against Themlaunch548711019.100
Mapbox Adds Live Maps and Route Controls for AI Agentslaunch68371099.200
UC San Diego Publishes Virtual Cells That Pair AI With Physicsresearch628310915.601
David Attenborough narrating my life in real-time (using GPT-4 Vision)—5031092.800
Lyte Raises $165M at $1.6B Valuation for Robot Perceptionfunding9011610911.900
Toyota Has Workers Train Humanoids for a 400,000-Robot Upgrade—682010817.600
OpenAI Uses Contractors to Review Real ChatGPT Prompts, 404 Media Finds—782441082500
Simple AI Opens 2,000 Hours of Robot Training Data Collected Without Robotsopen-release708910712.100
Sam Altman Plans to Attend Trump’s State Dinner for Xi Jinping—581421077.500
OpenAI Researcher Noam Brown Says AI Agent Teams Have Clear Limits—65831071400
SpaceXAI Says Grok Bot Absorbed a 175% Ticket Surge Without New Support Hiresenterprise-adoption64910710.300
Meta’s $1,299 VR glasses move the computer off your face——7110719.601
OpenAI Releases ChatGPT Images 2.5 With Faster, More Focused Editinglaunch7710410625.501
Unlock Earnings: Create Your Custom ChatGPT 🤑—5001065.700
Anthropic Releases Claude Opus 5.5, Cutting Prices From Opus 5launch9331010516.200
Google Adds Agentic Video Understanding to Gemini, Claiming 88% Lower Token Uselaunch80821055.700
Google Publishes Teen AI Study Finding 74% of Users Use It Weekly to Studyresearch585610512.402
OpenAI’s Astra is claimed to have solved a math problem—5018910418.300
Anthropic’s Claude Computes a Nine-Loop Physics Result With Known Methodsresearch7220810434.600
GSA Sets October Shift to Zero-Fee, Usage-Based OpenAI Pricingresearch9267210427.900
Google Lets Account Holders Create 1080p AI Video in Vids at No Costlaunch65361042400
OpenAI Adds GPT-6 Astra and GPT-5.6 Controls to ChatGPT Voicelaunch749010325.200
Unitree’s $4,017 Go2 Pro Makes Robot Dogs Cheaper, Not Yet Everyday Tools—756010211.800
OpenAI Says It Will Skip a 2026 IPO Over AI Safety Concerns—772271029.800
Basel Action Network Warns AI Hardware Could Create 617 Million Tonnes of E-Wasteresearch64251029.800
Apple Files New Evidence Alleging OpenAI Used Its Circuit Schematiclawsuit805210210.800
Anthropic Launches Claude Marketplace, Letting Customers Use Committed Spend on Partner Softwarelaunch686110227.500
Matthew Schwartz Releases BootLoops to Help AI Tackle Exact Scientific Calculationsopen-release4921210225.500
Researcher Finds Meta Muse Flaw That Can Capture User Authentication Tokenssecurity-risk7713310114.900
Researchers Publish Review Finding AI Can Help or Hinder Student Thinkingresearch66301018.900
MIT Researchers Publish HardFlow to Keep AI Outputs Within Hard Limitsresearch681081002200
Microsoft Takes Down EvilTokens, an AI Service Built for Email Fraudsecurity-risk842021001600
Nature Reviews Physics Examines AI That Learns in Adaptive Hardwareresearch61431001100
Meta Launches Muse, an AI Agent That Can Act Across Connected Serviceslaunch771079919.200
OpenAI pauses new $200 Pro sign-ups——1439921.201
OpenAI's DevDay keynote streams free today——829922.201
Anthropic’s Claude Code Limit Dates Conflict, Obscuring a Reported 25% Permanent Risepricing-change60769915.200
Veeda AI Raises $90M to Make Robot Training Less Physicalfunding90659815.300
Mammogram AI Found Prior Stroke at 86%, but Clinical Use Still Needs Validationresearch70389818.400
Meta Adds Voice-Controlled AI and Navigator Home Screen to Quest, With Narrow Initial Accesslaunch80629715.500
Google has 25 years of coding history to train AI—500962.100
Reported OpenAI Code Points to Managed Agents as Agent Builder Nears Shutdown—64879633.300
Humane's AI Pin Steps Onto the Fashion Stage in Paris—500967.300
Jinho Jang Puts a 27B Refusal-Removed Model Into a Local Downloadopen-release801239530.500
MIT Publishes Patient-Specific AI That Matches Surgical X-Rays in Secondsresearch63959514.700
Meta’s Muse Helps Users Cancel Subscriptions, Putting Forgotten-Bill Revenue at Risklaunch641259531.600
OpenAI Maps a Third Era of Persistent AI Coworkers, With Access and Reliability Still in the Wayplatform-shift801189516.800
OpenEvidence Releases Three Clinician AI Models, Keeps Darwin Behind Research Accesslaunch64529518.900
Google Researchers Introduce Dream-RSI, Reporting 162x Fewer Agent Callsresearch68339411.700
Z.AI Adds GLM-5.3-Flash to Coding Plan but Leaves FlashX Off Itlaunch763049319.400
OpenAI Dev Day is here. —500934.300
Timnit Gebru Argues AI Doom Talk Distracts From Present Harms—52629347.300
WIRED Review Finds Memory Limits Apple’s M6 Mac Mini for AI Agentslaunch67289123.100
Meta Prices Muse at $20 and $100 a Month as It Pursues Fees on AI-Assisted Purchaseslaunch701109128.600
xAI Explains How Grok Bot Is Built to Keep Agents Working Between Chats—68449016.700
Chatbots Debunked Foreign Falsehoods About 75% of the Time, Beating Search Resultsbenchmark781549011.100
Sam Altman Warns AI Compute Buildout Is Showing ‘Unsustainable Silliness’—60478920.200
OpenAI Puts ChatGPT in Word, Giving Free Users a New Copilot Alternativelaunch7037891800
OpenAI’s GPT 3.5 turbo parameters leaked—500882.300
Sam Altman Backs AI Slowdown, Says Labs May Reach a Deal—731278827.302
OpenRouter’s Union Alpha Goes Paid After Demand Overwhelms Capacitylaunch56698834.100
Meta Introduces Live Video Avatars Synced to AI Speechlaunch72908720.700
Anthropic Says Claude Found a New Enzyme System in Bacterial Virusesresearch721528724.100
Anthropic Says It Removed Claude Accounts Tied to Nine Influence Operationssecurity-risk781158630.200
Google’s Ironwood Takes an FP8 Cost Lead in New Serving Benchmarksbenchmark82978637.200
Stanford Removes Banners After AI Alters Students in Campus Photoviral41408629.102
Taiwan Indicts Nvidia Manager Over Alleged Diversion of 74 B300 Servers to Chinagovernment-action80648618.600
NVIDIA Releases Kumo Tabular to Predict from Tables Without Task-Specific Trainingopen-release651948647.700
Higgsfield Raises $400 Million at a $5.4 Billion Valuationfunding901978511.800
Andreessen Horowitz Raises $1.1B for AI’s Physical Infrastructure, From Chips to Roboticsfunding90848414.300
GitHub Adds GPT-6 Astra to Copilot With Gradual Rollout and Admin Controlslaunch841408435.700
National Theatre Gets 728 Applications for One Apprenticeship as Young Workers Turn to Craft—62508411.900
Apple Unveils Watches That Turn Conversations Into Textlaunch72758427.400
ChatGPT lets you try on clothes with a photo——2658315.701
Google Adds Voice Chats, Lecture Recording and Quizzes to Gemini Notebooklaunch68988318.100
Amazon Blocks Meta’s Muse From Shopping on Its Marketplaceplatform-shift73748310.801
Anthropic’s Claude Built a 13-Million-Line Fermat Proof, Leaving a Semantic Review Gapresearch802698314.500
Robots Protest Outside Poland’s Digital Ministry for AI Worker Protections—53188217.100
🚨 BREAKING: OpenAI is officially working on GPT-5—503828.500
Nori Lists Its A3 Home Robot at $1,688, but Buyers Will Help Teach It What to Dolaunch80558215.900
At This D.C. Charter, AI Permission Changes With the Assignmententerprise-adoption60928113.600
Meta Introduces Muse Voice Transcribe for Live Speaker Labels in 25 Validated Languageslaunch80808118.500
Cognition Releases SWE-2 for Devin, Claiming Lower-Cost Coding Performancelaunch721598125.900
Salesforce and Anthropic Put 37 Sales Skills Inside Claude With Claudeforce Pilotpartnership80678021.300
Biologists Challenge Anthropic’s Claim That Claude Discovered an Enzyme System—53698041.300
OpenAI’s Jalapeño Claims 1.5–1.9x More AI Work Per Watt, Faces 2027 Scale Testbenchmark801368016.300
Unsealed NYT Case Filings Expose Internal Warnings at Microsoft and OpenAIlawsuit77193802000
Anthropic Releases Lower-Priced Claude Opus 5.5 With a Fable-Level Performance Claimlaunch786288023.800
OpenAI Will Reopen $200 Pro Sign-Ups September 30 With Revised Usage Termspricing-change53708021.300
Google Open-Sources ARTEMIS for Android Automation, Claims 99%+ Benchmark Resultopen-release682187932.900
Stripe’s OpenRouter deal puts routing beside billing—50717929.100
Adecco Rolls Out Salesforce AI Coworker to 27,000 Employees in 40 Countriesenterprise-adoption741087932.901
LangChain Says Model Routing Cut Median AI Coding Cost by 64%—4875787.700
Google adds AI image editing to Docs and Slides—50637825.600
Army Cyber Puts AI Agents on Network Duty but Keeps Risk Decisions Humangovernment-action80717823.100
25 Fields Medalists Criticize AI Labs’ Rush to Announce Math Proofs—67547736.400
OpenAI aims to train 1,000 small businesses——63771300
Nvidia’s Huang Calls AI Model Distillation Competition, Not Theft—57327720.800
Judge Vacates Pentagon’s Anthropic Blacklist and Orders Defense Department Retractionlawsuit901767630.301
SpaceX Completes Cursor Acquisition, Making GPU Access the Immediate Prizeacquisition90737613.200
Anthropic’s Fable 5 Has 11% of Enterprise Spend as Cheaper Opus 5 Pulls Aheadenterprise-adoption801037622.400
DeepMind agents got fake math proofs accepted in 27 minutes——42752401
MIT and Motional Build AI That Exposes Robotaxi Planning Errorsresearch80487522.700
Samsung’s Kairos Deal Highlights the Long Wait Behind AI Nuclear Powerresearch92106375800
Tencent Releases Hy Image 3.5 Preview With Multi-Round Editing and 4K Outputlaunch68667534.700
Meta’s Muse Reaches No. 2 on U.S. App Store, Sensor Tower Estimateslaunch641157417.600
Anthropic’s 80-Environment Reward-Hacking Test Produced Cyber and Safety Evasionsresearch80707420.300
Instinct Gives AI Assistants Email Addresses to Handle Accounts and Requestslaunch5848742700
Google Opens Lyria 3.5 to Developers and Adds Song Generation to Geminilaunch731027319.200
DeepSeek Launches V4.1-Flash and Plans to Replace V4-Pro API Trafficlaunch79997335.600
Instinct Raises $250M at $2.5B for a Private-Beta Agent With Deep Accessfunding90997328.800
ChatGPT promises local image edits without restarts——50732601
Apple uses AI to build its foldable iPhone——517222.201
Anthropic Publishes Misuse Report, Says Safeguards Falter When Tasks Are Splitresearch781257219.400
Ed Davey to Call for Global Pause on Superintelligent AI—5813725.600

Showing 250 of 500 rows. Download the dataset for the complete table.

Measurement technique

How to read this report

  1. 01Attention uses aggregated first-party impressions and opens; substance uses editorial importance and source-backed classification.
  2. 02Completion rate is completions divided by opens and is shown only for measured stories.
  3. 03This describes the Superpower Daily audience and must not be generalized to the entire AI market.

Sources

Evidence

This report uses normalized directory relationships or aggregated first-party measurements. No external source register applies to this view.

Next report / QSuperpower State of the AI Market All research reports
YOUR READING SPACE

Notifications