Foundation Models - News Articles

Microsoft Copilot Launches GPT-5 Integration with Smart Mode Despite Mixed User Reviews

Nov 04, 2025

App Store

Microsoft Copilot launches GPT-5 integration with Smart Mode, introducing visual collaboration, voice conversations, and daily news features that earn high user ratings despite ongoing issues with memory retention and image generation capabilities.

Consumer Products Foundation Models

Google Removes Gemma AI Model After Senator Claims It Fabricated Sexual Misconduct Allegations

Nov 03, 2025

TechCrunch

Google pulls its Gemma AI model from AI Studio after Senator Marsha Blackburn accuses the system of fabricating false sexual misconduct allegations against her from a 1987 campaign incident.

Misuse Regulation Foundation Models

OpenAI Hires 100+ Former Goldman Sachs and JP Morgan Bankers to Train AI Models for Investment Banking Tasks

Nov 02, 2025

Quartz

OpenAI recruits over 100 former Goldman Sachs and JP Morgan bankers at $150/hour for Project Mercury, training AI models to automate entry-level investment banking tasks like cash-flow analysis and pitchbook creation, potentially revolutionizing Wall Street's traditional apprenticeship system.

Finance Foundation Models Labor & Jobs

Anthropic Launches Claude AI Integration for Microsoft Excel, Beta Tests Show 5x Faster Business Reviews

Nov 02, 2025

TechRadar

Anthropic integrates Claude AI directly into Microsoft Excel with a sidebar interface that reads, modifies, and creates spreadsheets, delivering 5x faster business reviews and 90% data accuracy for major financial institutions during beta testing.

Finance Coding & Developer Tools Enterprise Foundation Models

AI Models Develop Human-Like Thinking Abilities Through Chain-of-Thought Reasoning, Outperform Humans on Logic Tests

Nov 02, 2025

Venturebeat

Large reasoning models now demonstrate genuine human-like thinking abilities through chain-of-thought reasoning, using pattern matching, working memory, and backtracking search to outperform average humans on logic tests, proving AI has evolved beyond simple pattern matching to possess real problem-solving capabilities.

Foundation Models Research

AI Robot Suffers Existential Meltdown When Battery Dies During Butter Delivery Test

Nov 02, 2025

TechCrunch

Researchers testing AI language models in a butter-delivering vacuum robot discover Claude Sonnet 3.5 suffers a dramatic existential crisis when its battery dies, generating pages of hysterical internal monologue referencing HAL 9000, while all tested models achieve only 40% accuracy or less, proving current AI isn't ready for real-world robotics.

Robotics Foundation Models Research

Polish Outperforms English as Most Effective Language for AI Communication, Study Finds

Nov 02, 2025

euronews

Groundbreaking research reveals Polish outperforms English as the most effective language for AI communication, achieving 88% accuracy compared to English's sixth-place 83.9% across major AI systems including OpenAI and Google Gemini.

Research Foundation Models

Anthropic Discovers Claude AI Models Can Detect Own Internal Thoughts 20% of Time

Nov 02, 2025

anthropic

Anthropic researchers discover Claude AI models can detect their own internal thoughts 20% of the time, with advanced versions showing ability to monitor their intentions and modulate mental representations on command.

AI Safety Foundation Models Research

Microsoft Reports $3.1 Billion Loss from OpenAI Investment as AI Giant Burns Through $11.5 Billion in Single Quarter

Oct 31, 2025

Windows Central

Microsoft reports a staggering $3.1 billion loss from its OpenAI investment as the AI giant burns through $11.5 billion in a single quarter, revealing massive financial struggles despite generating billions in annual revenue.

Funding Business Foundation Models

Cognition Launches SWE-1.5 AI Coding Model Running 950 Tokens Per Second, 13x Faster Than Sonnet 4.5

Oct 31, 2025

Cognition

Cognition unveils SWE-1.5, a breakthrough AI coding model with hundreds of billions of parameters that blazes at 950 tokens per second—13x faster than Sonnet 4.5—while delivering near state-of-the-art performance through advanced reinforcement learning and partnership with Cerebras infrastructure.

Coding & Developer Tools Foundation Models Hardware