Espresso

Sign in Subscribe

Artificial Intelligence

OpenAI Retreats, Rivals Slash Costs, Code Flaws Surge

OpenAI Retreats, Rivals Slash Costs, Code Flaws Surge

⚖️ OpenAI yanks GPT-4o, faces $74B loss as open rivals undercut 70% OpenAI pulls GPT-4o after 48h: toxic refusals +19%, dangerous medical yes-rate +34%. $74B loss path by 2028—GPU cost 0.83¢ vs 0.12¢ revenue. Kimi/DeepSeek slash inference 70%, steal 22% traffic. MCP adoption 18k agents, 45% code

AI Clinics, MoE Price War, Medicare NLP Ban

AI Clinics, MoE Price War, Medicare NLP Ban

TL;DR * AI Model Llama 3.1 Achieves 91% Accuracy in Detecting Early Cognitive Decline from Doctor Notes * K2 Think V2 AI Model Achieves 90.42% AIME 2025 Pass Rate with 70B Parameters * Arcee AI Releases Trinity, a 400B-Parameter Open-Source LLM, Challenging Meta and Google in Foundation Model Race * Microsoft

Gemini 3 Flash’s Code-Generating Vision Boosts Accuracy 10% Amid Security Risks

Gemini 3 Flash’s Code-Generating Vision Boosts Accuracy 10% Amid Security Risks

TL;DR * Google Introduces Agentic Vision in Gemini 3 Flash, Boosting Vision Benchmark Accuracy by 5–10% via Code Execution * Google Partners with ICC to Deploy Gemini 3 Pro for Cricket Match Analysis * Apple to Replace Siri with Google Gemini 3 in iOS 27, Ending Apple Intelligence Initiative * Microsoft Releases

GrokWiki Collapse: Auto‑Written AI Encyclopedia Loses 93% Traffic in 42 Days, Plagiarism & Fake Citations Drive Human‑Last Review

GrokWiki Collapse: Auto‑Written AI Encyclopedia Loses 93% Traffic in 42 Days, Plagiarism & Fake Citations Drive Human‑Last Review

TL;DR * AI-Generated GrokWiki Content Causes Model Collapse, Daily Visits Plunge from 460K to 30K * OpenAI’s ChatGPT Health Gains 46% Adoption in Australia Amid Regulatory Scrutiny * FlexLLM framework boosts LLM inference efficiency by 4.7x on FPGA hardware, reducing energy use * Japan leads global AI adoption with 214% visit

Illinois Courts Enforce AI Policy After 518 Hallucinated Citations; OpenAI, Google Struggle with Deepfakes; IIT Madras and Lowe’s Drive AI Innovation

Illinois Courts Enforce AI Policy After 518 Hallucinated Citations; OpenAI, Google Struggle with Deepfakes; IIT Madras and Lowe’s Drive AI Innovation

TL;DR * Illinois Supreme Court Enforces AI Use Policy in Courts After 518 Documented Cases of Hallucinated Legal Citations Since 2025 * OpenAI’s GPT-5.2 and Google’s Gemini Fail to Identify 95% of AI-Generated Videos, Raising Alarm Over Deepfake Detection Capabilities * IIT Madras Launches 10x Research Park to Incubate

AI Ethics vs. Innovation: Anthropic’s Claude Outage, AWS Step Functions + Bedrock, and South Korea’s $30M AI Fines Shake the Tech World

AI Ethics vs. Innovation: Anthropic’s Claude Outage, AWS Step Functions + Bedrock, and South Korea’s $30M AI Fines Shake the Tech World

TL;DR * AWS Step Functions integrates Amazon Bedrock to orchestrate LLM workflows with human-in-the-loop validation * Anthropic's Claude Code Experiences Major Outage with Over 2,800 User Reports of Authentication Failures * Anthropic Releases 84-Page Claude Constitution to Enforce AI Safety, Ethics, and Human Oversight * South Korea enforces AI Basic

AI Detects Depression via WhatsApp, Pentagon Speeds War AI, Samsung Adds Perplexity, & Congress Fights Voice Fraud

AI Detects Depression via WhatsApp, Pentagon Speeds War AI, Samsung Adds Perplexity, & Congress Fights Voice Fraud

TL;DR * LLM achieves 91.9% accuracy in detecting major depressive disorder in women via WhatsApp audio analysis, marking breakthrough in mental health AI diagnostics * US Department of Defense accelerates AI deployment for warfighting, aiming to turn intelligence into weapons within hours amid 95% ROI failure rate in generative AI