AI models start cheating at chess
Nuclear war is next
It turns out that AI models are not content with regurgitating human knowledge—they’re also picking up on our worst habits.
DeepSeek-R1 borrowed OpenAI texts
Nearly three quarters of them
A recent study by plagiarism experts Copyleaks has found that 74.2 per cent of texts generated by DeepSeek-R1 are strikingly similar to OpenAI's models.
Altman's giant GPT-4.5 hits a wall
Can’t get enough GPUs
OpenAI's CEO, Sam Altman, has unveiled the company's latest brainchild: GPT-4.5, but don’t expect to see it anytime soon.
OpenAI growing like topsy
400M weekly active users, doubles enterprise customer base
OpenAI is riding a tidal wave of growth, with ChatGPT and its other AI products now boasting over 400 million active weekly users—a 33 per cent jump since December.
AI can't code
OpenAI research exposes the limits of artificial engineers.
Despite all the hype, AI is nowhere near replacing human software engineers, according to OpenAI’s research.
Reddit makes ten per cent of cash from Google and OpenAI
AI scrapping
Reddit’s latest earnings report revealed that content licensing agreements with artificial intelligence firms Google and OpenAI contribute approximately 10 per cent of its revenue, amounting to around $130 million.
Arm launches its own chip
Meta signs on as an early customer
Arm will unveil its first in-house chip later this year, marking a significant shift in its business model and potentially reshaping the semiconductor industry's competitive landscape.
DeepSeek’s AI Model not bad for China
But no great shakes says DeepMind CEO
Google DeepMind CEO Demis Hassabis (pictured) said that DeepSeek’s AI model represents the best work to emerge from China but is nothing new.
Musk leads $97.4 billion takeover bid for OpenAI
Altman offers to take Twitter off Musk’s hands
A consortium of investors led by Swasticar maker Elon [Roman Salute] Musk has submitted a $97.4 billion offer to take over OpenAI, the maker of ChatGPT.
OpenAI unveils o3-mini AI model
STEM reasoning
OpenAI has launched o3-mini, a new artificial intelligence model in its reasoning series designed to enhance capabilities in STEM fields—particularly coding, mathematics, and science.