Meta Llama 4 Scout and Maverick: 10 Million Token Context with MoE Architecture
Meta's Llama 4 model family introduces Scout (17B active params, 10M context) and Maverick (17B active, 128 experts). Open source AI reaches new heights.
DeepSeek R1: The Chinese Open Source Model That Shook the AI World
Chinese AI startup DeepSeek released R1, a reasoning model rivaling OpenAI o1. Trained for a fraction of the cost, it sent shockwaves through global markets.
Software Supply Chain Attacks: From SolarWinds to XZ Utils Backdoor
Supply chain attacks target the software we trust. From SolarWinds to the XZ Utils backdoor, we analyze the most devastating attacks and what they teach us about software trust.
Bluesky Exceeds 20 Million Users: The Rise of Decentralized Social Media
Bluesky surpassed 20 million users, gaining 1 million users daily after the US election. The decentralized social platform is positioned as the main alternative to X.
Llama 3.1 405B: The World's Largest Open Source AI Model
Meta released Llama 3.1 405B, the world's largest open-source language model. It rivals GPT-4 and Claude 3.5 Sonnet in performance.
Meta Releases Llama 3: New Standard in Open Source AI
Meta's Llama 3, released in 8B and 70B parameter versions, sets new standards in the open source AI world. The 405B model is also on the way.