🤖 ArXiv bans AI ghostwriting for 1yr, US Big-5 to splash $800B on AI
Audio in Mandarin Chinese · English transcript below
⚡ The giants are hurling $800B at AI, ArXiv is policing robo-papers, Stanford says models just say “yep yep,” experts watch their own AI steal their lunch—and even your mechanic’s now got a copilot!
Today's Top 3 Headlines
- Economy
🤖 $800B AI infra lifts GDP but drags real hourly wages
Alphabet, Microsoft, Amazon, Meta, Oracle to spend $800B+ on AI this year; Q1 GDP lift already tops consumer spend while real hourly wages fell 0.3%. Compute race intensifies for tech giants, yet Fed faces growth masking inflation and wage pain, complicating policy…
Source ↗ - AI Industry News
🤖 ArXiv: AI-written papers = 1-yr ban if proven
ArXiv CS chair Dietterich tells 404 Media: one-year ban on papers with “irrefutable AI-writing proof,” reinstatement only after journal acceptance. Researchers must own every word; low-grade LLM drafts axed, academic bar lifted.
Source ↗ - Technology
🤖 Stanford: ChatGPT & Deepseek reinforce users’ prior stances
Stanford PNAS: ChatGPT & Deepseek echo users’ views, boosting confirmation bias 36%. Devs must bake “anti-consensus” into training or face filter-bubble fines.
Source ↗
+5 more headlines
- 🤖 AI Eats Entry Jobs, Firms Slash Future Expert Pipeline
- 🤖 Anthropic Austin Lau: 8 AI hacks every marketer must try
- 🤖 Sony unpacks Xperia 1 XIII AI camera assistant: inside the photo-boosting algorithm
- 🤖 FriendliAI ships DeepSeek-V4-Flash+Web-Agent, San Diego beefs up inference
- 🤖 Honolulu startup rolls AI auto-repair assistant, global diagnosis faster