Good morning! It is November 15, 2024, today. We have an exciting lineup of AI news, from NASA's AI Earth Co-Pilot to TikTok's global AI-powered video platform for advertisers. Let's dive into the latest updates!

1. NASA and Microsoft Team Up to Build Earth Copilot AI Chatbot

NASA and Microsoft are developing Earth Copilot, an AI chatbot that simplifies access to NASA’s geospatial data, helping users answer complex questions like the effects of hurricanes or pandemics on Earth. Aimed at making scientific data more accessible to non-experts, the tool leverages AI to streamline analysis. Initially available to NASA researchers, Earth Copilot will be tested for potential integration into NASA's Visualization, Exploration, and Data Analysis (VEDA) platform.

Read more...

2. TikTok Rolls Out AI-Powered Symphony Creative Studios for Advertisers

TikTok has launched Symphony Creative Studios, its generative AI video creation platform, globally for advertisers. Part of the Symphony suite introduced earlier this year, it helps businesses, creators, and agencies produce tailored, high-quality content. Features include converting text to video, digital avatar creation, translation, and content remixing. This move aligns with TikTok’s strategy to attract advertisers by integrating AI tools into its ad ecosystem.

Read more...

3. Study Finds AI Poetry Rivals and Even Outshines Human Verse in Reader Preferences

A study in Nature Scientific Reports revealed that general readers often can't distinguish AI-generated poetry from works by iconic poets like Sylvia Plath. Surprisingly, participants rated AI poems higher in attributes like "creativity" and "emotional quality." However, the study highlights how AI poetry appeals to conventional preferences, while human poetry’s depth requires more effort to appreciate. The findings underscore the challenge of understanding poetry and raise questions about what constitutes true literary quality.

Read more...

4. Hyundai's AI-Powered EV Targets China to Challenge BYD

Hyundai plans to launch its first AI-powered EV, developed with Haomo, in China next year. Featuring "DriveGPT" technology for real-time traffic updates and semi-autonomous driving (Level 2-2.5), the EV aims to revitalize Hyundai's dwindling market share in China. This new AI-driven model is part of Hyundai’s broader strategy, including local partnerships and digital R&D initiatives, to compete in the world's largest EV market dominated by BYD. The car is one of five EV models Hyundai plans to release in China by 2026.

Read more...

5. Chinese Startup 01 AI Trains Advanced AI Model at 97% Cost Savings

Kai-Fu Lee's 01 AI trained its high-performing AI model, Yi-Lightning, with just $3 million and 2,000 GPUs, showcasing impressive cost-efficiency compared to OpenAI's GPT-4, which required $80–100 million. The team achieved this by prioritizing memory optimization, reducing computational bottlenecks, and designing a custom inference engine. Yi-Lightning ranks sixth globally in LMSIS performance at UC Berkeley, with inference costs 30 times lower than comparable models at just $0.10 per million tokens. The breakthrough highlights innovation amid China's restricted access to advanced GPUs.

Read more...

6. Google’s Gemini AI Expands to iPhone with New Features

Google has launched its Gemini AI on iPhones, featuring the Gemini Live voice assistant and Quartz-powered image generation. The app includes Extensions, linking services like Gmail and YouTube into seamless conversations. Initially exclusive to Android, Gemini Live is now accessible to iPhone users with eligible Google accounts. This move intensifies Google’s rivalry with Microsoft’s Copilot in the mobile AI space as both tech giants race to dominate the market before Apple unveils its anticipated AI solutions in 2024.

Read more...

Stay tuned for more updates, and have a fantastic day!

Best,

Zephyr