🌏
News
OpenAI Offers $250K Prize for Discovering Lost Amazonian Cities Using Its Models
OpenAI has launched the OpenAI to Z Challenge, offering a $250,000 prize to anyone who can use its models to help find ancient, lost cities in the Amazon. The goal is to use AI to analyze satellite images and discover signs of past civilizations hidden under forests. The challenge is open to researchers, scientists, and curious minds around the world.
Nvidia, Cisco, Oracle and OpenAI are backing the UAE Stargate data center project
During his recent Middle East tour, President Trump announced major AI deals with Gulf nations, including allowing Nvidia to sell 500,000 AI chips annually to the UAE and 18,000 to Saudi Arabia.
Microsoft and Hugging Face expand collaboration to make open models easy to use on Azure
Hugging Face and Microsoft have teamed up to launch Azure AI Foundry, a new platform that makes it easier for companies to build and manage their own AI models. It offers tools to train, fine-tune, and deploy open models securely within Azure. Users can pick top open-source models and customize them without needing deep AI knowledge. It’s designed to help businesses move faster with AI while keeping control of their data.
Databricks is acquiring Neon, a startup that built a fast, cloud-native version of Postgres designed for developers and AI agents. Neon’s serverless architecture lets users create and scale databases instantly, with features like branching and forking—much like Git for code. Notably, over 80% of Neon’s databases are now created by AI agents, showing how AI-driven development is growing fast.
23andMe and its user data will soon belong to a pharmaceutical giant
23andMe is selling its drug research arm to biotech company Regeneron for $25 million after filing for bankruptcy. This move comes as 23andMe struggles with falling stock prices and privacy concerns. The deal gives Regeneron access to genetic data that was once a key part of 23andMe’s business.
The models are getting better, but they’re also more likely to be good at bad stuff
At a recent White House summit, major tech companies—including Meta, Google, and OpenAI—pledged to follow voluntary AI safety guidelines. These include testing AI systems for security risks, adding watermarks to AI-generated content, and being open about their tools’ strengths and weaknesses. While these steps aim to make AI safer, critics note that without legal enforcement, the promises may not be enough.
AI scam factories force trafficked workers to defraud global victims
Young Indonesians are being lured by fake tech job ads on Facebook and Telegram, only to end up trafficked into scam centers across Southeast Asia. These victims are forced to use AI tools like deepfakes and voice clones to deceive people worldwide, often targeting them through social media and dating apps. Many have their passports taken and are trapped under strict control, working in harsh conditions. This growing cybercrime has led to billions in global losses, with Americans alone losing $12.5 billion in 2024.
A new cold war is brewing over rare earth minerals
China has implemented new export controls for rare earth minerals and magnets. The changes could upend the shift to electric vehicles.
🤯
Reads
Microsoft CTO Kevin Scott on how AI can save the web, not destroy it
In a recent interview on The Verge’s Decoder podcast, Microsoft CTO Kevin Scott discussed the company’s new open-source tool that allows websites to add AI-powered natural language search easily. This tool uses the Model Context Protocol (MCP), enabling sites to control their data and AI interactions without relying on major search engines like Google or Bing. Scott believes this approach can make the web more open and fair, giving website owners more control and promoting innovation. He also emphasized the need for new business models that support content creators and ensure ethical use of AI.