The Latest Breaking News from Scale AI

OpenAI’s o1 and DeepSeek’s R1 models, which previously sat atop the leaderboard, could only get through roughly 9% of the exam.

A groundbreaking AI benchmark called Humanity's Last Exam looks to test LLM's reasoning capabilities. Let's just hope no model ever passes the test…

Meet the serial entrepreneur who rides her electric skateboard to work—when her assistant isn’t driving her.

Dan Hendrycks, AI safety advisor to Elon Musk's xAI, recently posted on his LinkedIn profile that he has taken on a similar role at Scale AI.

While investors debate the business potential of AI, some eager early customers appear to be U.S. defense and security agencies.

In dusty factories, cramped internet cafes and makeshift home offices around the world, millions of people sit at computers tediously labelling data.

Two of San Francisco’s leading players in artificial intelligence have challenged the public to come up with questions capable of testing the capabilities of large language models (LLMs) like Google Gemini and OpenAI’s o1. Scale AI, which specialises in preparing the vast tracts of data on which the LLMs are…

Arizona Iced Tea stays true to its 99-cent price and big cans by producing faster, shipping better, shipping closer, and using thinner cans.

Scale AI, a hot startup, prioritises merit and predicts AI skills will be crucial for future job opportunities.

Scale AI, which helps companies label and test data for AI model training, has closed a new $1 billion funding round at a $14 billion valuation.

LLMs may be getting better at passing benchmark tests by 'memorizing' the answers