AI News

AI startup Sieve releases an API that corrects eye contact in videos for just 10¢ per minute, automatically adjusting gaze in real-time to make content more engaging.

Google.org commits $15 million to train government employees in AI, supporting the U.S. Department of Energy’s push for an AI-ready workforce.

Germany’s education chief Christine Streichert-Clivot urges the systematic integration of AI in classrooms, promoting modernization while upholding ethical standards.

Endor Labs has launched "Endor Scores for AI Models," a tool that rates AI models based on security, popularity, quality, and activity. This innovation helps users quickly identify the most secure open-source AI models on Hugging Face, enhancing ease of selection and trust within the AI community.

BharatGen launches e-vikrAI, a vision-language model for Indic e-commerce. Sellers can upload product images to receive auto-generated titles, descriptions, features, and pricing tailored with cultural insights.

Ola co-founder Ankit Bhati and others invest $1 M in Tablesprint, an AI no-code platform that lets businesses quickly create custom applications.

In August 2024, researchers identified that nearly 4.36% of 2,909 new English Wikipedia articles featured substantial AI-generated content.

Lenovo introduces AI Buddy, a smart digital assistant that seamlessly manages your schedule, reminders, and daily tasks. Elevate your productivity with this efficient AI companion.

Lenovo is FIFA’s Technology Partner for the 2026 World Cup, which will feature 48 teams across three host nations. They’ll also support the first Women’s World Cup in South America, held in Brazil in 2027.

OpenAI finds ChatGPT exhibits significantly less gender bias than humans. While user-chosen usernames can slightly influence responses, the effect is minimal and mostly limited to older or less aligned models.

The New York Times is suing AI startup Perplexity, accusing it of unlawfully using their content to fuel its AI-powered search engine.

Scientists at Meta, UC Berkeley, and NYU have created "Thought Preference Optimization," a technique that encourages AI models to deliberate more carefully before responding, enhancing their performance.