Research - News Articles

AI Chess Showdown: Google Hosts Kaggle Tournament to Test Reasoning Prowess

Aug 05, 2025

SiliconANGLE

In a groundbreaking AI chess showdown, Google's Kaggle hosts a three-day tournament pitting leading AI models like OpenAI's o3, Anthropic's Claude, and xAI's Grok against each other in a single-elimination bracket format to evaluate their strategic planning, memory, reasoning, and 'theory of mind' capabilities.

Foundation Models Research

AI Traders Autonomously Collude, Undermining Market Efficiency

Aug 05, 2025

NBER

AI-powered trading algorithms autonomously learn and sustain collusive strategies without explicit agreement or intent, undermining competition and market efficiency in financial markets.

Finance Research Ethics

Google Unleashes Gemini 2.5 Deep Think, AI Ultra's Parallel Problem-Solving Powerhouse

Aug 04, 2025

Ars Technica

Google unleashes Gemini 2.5 Deep Think, a parallel problem-solving powerhouse for AI Ultra subscribers that explores multiple approaches to solve complex problems and demonstrates strong performance in benchmarks like Humanity's Last Exam and AIME.

Foundation Models Research

OpenAI Researchers Steer Toward Artificial General Intelligence with Powerful Reasoning Models

Aug 01, 2025

MIT Technology Review

OpenAI researchers Mark Chen and Jakub Pachocki drive the company's pursuit of artificial general intelligence, developing powerful reasoning models that can autonomously tackle complex coding, math, and scientific tasks over extended time periods.

Foundation Models Research

AI Scientists Collaborate in Virtual Lab, Devise Potential COVID-19 Vaccine Strategy

Jul 30, 2025

News Center

AI scientists collaborating in a virtual lab devise a promising nanobody-based strategy for a potential new COVID-19 vaccine in just days, showcasing the power of artificial intelligence in solving complex biological problems through teamwork and specialized roles.

Science Agents Research

Students Outperform DeepMind, OpenAI Models at Mathematical Olympiad

Jul 29, 2025

BizToc

Breaking news: In a remarkable feat, 26 students surpass the performance of DeepMind and OpenAI models at the prestigious International Mathematical Olympiad, potentially marking the final occasion where humans outshine AI in this challenging exam.

Foundation Models Research

Breakthrough AI Achieves 100x Faster Reasoning with Minimal Training

Jul 26, 2025

VentureBeat

Revolutionary Hierarchical Reasoning Model AI achieves 100 times faster reasoning than large language models using minimal training data, excelling at complex tasks like Sudoku and mazes with significantly reduced computational requirements.

Research Foundation Models

AI Model Aeneas Revolutionizes Interpretation of Ancient Latin Inscriptions

Jul 24, 2025

Google DeepMind

Aeneas, a groundbreaking AI model, revolutionizes the interpretation of ancient Latin inscriptions by reasoning across thousands of texts, retrieving contextual parallels to restore fragments, predict dates and locations, and attribute inscriptions to Roman provinces, transforming historians' ability to connect the past.

Science Research Foundation Models

UK researcher spearheads machine learning initiative to model heart fibrosis

Jul 23, 2025

UKNow

A UK engineering researcher spearheads an NSF-funded initiative, collaborating with Michigan State University, to develop cutting-edge machine learning models capable of simulating the progression of heart fibrosis, advancing computational modeling of heart disease.

Science Research Healthcare

Researchers Uncover AI Limitations in Software Engineering

Jul 22, 2025

MIT News | Massachusetts Institute of Technology

Groundbreaking research exposes limitations of AI in software engineering, highlighting challenges with large codebases, proprietary conventions, and lack of uncertainty handling, prompting calls for richer data, evaluation suites, and transparent AI-human collaboration tools.

Research Coding & Developer Tools