Daily AI NEWS - September 14, 2024
AI, Disinformation, and Responsibility

OpenAI has introduced its new o1 model, internally code-named Strawberry, which emphasizes reasoning and problem-solving capabilities. This model, available in ChatGPT and the API, is designed to handle complex tasks in areas like science and coding more effectively than its predecessors.

Impressively scoring on human intelligence benchmarks, the o1-preview model also poses substantial risks, with OpenAI acknowledging that the new models could be misused to create biological weapons. Despite these advances, the company has noted that the model can still provide incorrect answers.

Subscribers of Plus and Teams can already try the o1-preview model. However, safety concerns remain as detailed in the o1 safety card, which highlights issues such as scheming, reward hacking, and potential misuse in creating biological threats.

Right-wingers are spreading anti-Haitian AI disinformation, leading to threats against Springfield, Ohio’s Haitian community. Racist memes have escalated tensions and caused real-world harm.

Meta has admitted to scraping public Facebook and Instagram posts and photos of adults dating back to 2007 to train its AI models. Americans can't opt-out but have some alternatives.

A Ukrainian startup, Osavul, plans to defend companies from disinformation using AI. Founded in 2022, it combats Russian propaganda by analyzing threats through AI.

1minAI has launched a groundbreaking all-in-one platform integrating multiple advanced AI models, offering lifetime access for just $39.99.

TECNO is incorporating "practical AI" in its new foldables, targeting emerging markets to bridge the tech gap between developed and developing countries.

Google's Gemini Live AI feature, once exclusive to the Pixel 9, is now free for all Android devices, increasing accessibility and user engagement.

Demand for Nvidia's AI chips, including the Blackwell line, has surged, straining suppliers. OpenAI's Sam Altman revealed a close working relationship with the executive branch and numerous members of Congress. Meanwhile, OpenAI is reportedly raising at a $150 billion valuation, making it worth more than 88% of Fortune 500 firms, and looking to raise $5 billion in debt.

French AI startup Poolside, based in Paris, is in discussions to raise nearly $500 million at a $3 billion valuation. Poolside is developing a ChatGPT-like large language model that writes software code; specifics on their business plan remain undisclosed.

Hasbro’s CEO Chris Cocks believes the adoption of AI in D&D is inevitable, emphasizing that it’s a matter of when, rather than if.

The latest Fortnite missions for Chapter 5, Season 4, pit players against Marvel's Mysterio in Jones and Hope's ongoing battle plans, adding another layer of excitement.

CCP Games has announced EVE Frontier, a blockchain-based survival MMO set in the EVE Online universe, transforming Project Awakening into a new gaming frontier.