It’s time for Bits & Bytes…
… where we bring you news, innovations, and thought-provoking insights from AI, IT, and beyond. In this week’s newsletter we’re looking at:
-
Musk launches Grok3 claiming to outperform OpenAI.
-
DeepSeek faces restrictions in South Korea.
-
OpenAI is changing the way it trains it’s model.
IN THE SPOTLIGHT
🚀 xAI Launches Grok 3
Elon Musk’s xAI introduces Grok 3, powered by 200,000 GPUs and claiming to outperform OpenAI’s models on key benchmarks. The new AI family includes reasoning models, DeepSearch capabilities, and upcoming voice features, with access tiered between X Premium+ ($50/month) and a new SuperGrok subscription ($30/month).
Musk positions Grok 3 as a “maximally truth-seeking AI,” promising to open-source its predecessor while aiming for political neutrality in response to previous criticism.
TL;DR
-
Claims superior performance to GPT-4o.
-
Includes new reasoning capabilities.
-
Launches DeepSearch feature.
-
Introduces tiered pricing model.
TECH HEADLINES FROM ACROSS THE WEB
🔒 South Korea Blocks DeepSeek Over Privacy Concerns
South Korean officials have restricted access to DeepSeek’s app after discovering user data transfers to ByteDance, joining other countries like Australia and Italy in limiting the Chinese AI company’s reach. While existing users can continue using the service, authorities strongly advise against sharing personal information until privacy compliance is achieved.
🗞️ NYT Cautiously Adopts AI for Newsroom
The New York Times approves select AI tools for newsroom tasks like editing and summarization, while maintaining strict guidelines against AI-drafted articles. The move, including the launch of internal tool Echo, comes despite ongoing legal battles with OpenAI over training data rights.
🛠️ OpenAI Overhauls ChatGPT’s Training
OpenAI announces a significant policy change in its AI training, embracing “intellectual freedom” and committing to present multiple perspectives on controversial topics. The update comes amid Silicon Valley’s broader shift in AI content moderation, with ChatGPT now set to offer balanced viewpoints rather than avoiding sensitive subjects.
TECH FOR GOOD
🌟Geneva Pioneers AI-Driven Mental Healthcare Hub
Switzerland’s University Hospitals of Geneva partners with the Wyss Center to create a groundbreaking AI hub focused on treating neurological and psychiatric disorders. The 1,000m² facility will combine advanced AI technology with clinical care, offering patients access to cutting-edge treatments while developing new AI solutions for early disease detection and prevention.
