ChatGPT 4 - Search the web and tell me the latest news about OpenAI

Search the web and tell me the latest news about OpenAI.

ChatGPT

OpenAI has recently introduced several exciting developments across its product range, signaling continuous innovation in AI technologies. Here's a summary of the latest updates:

New Embedding Models and API Updates

OpenAI has released a new generation of embedding models, including the highly efficient text-embedding-3-small model and the more powerful text-embedding-3-large model. These models are designed to better understand the relationships between content, facilitating tasks like clustering or retrieval. OpenAI has also updated the GPT-4 Turbo and GPT-3.5 Turbo models, introduced new API usage management tools, and announced lower pricing for GPT-3.5 Turbo​ (OpenAI)​.

Developer Products Announced at DevDay

During the DevDay event, OpenAI announced improvements to GPT-4 Turbo's instruction-following abilities and introduced a JSON mode for generating valid JSON responses. Additionally, a new seed parameter for GPT-4 Turbo and GPT-3.5 Turbo allows for reproducible outputs, enhancing the development of features such as autocomplete in search experiences. The new Assistants API was also launched, enabling developers to create AI-powered apps with agent-like experiences. This API provides capabilities such as a Code Interpreter, Retrieval, and function calling​ (OpenAI)​.

Introduction of Sora, A Generative Video Model

OpenAI teased a new generative video model named Sora, capable of producing high-definition video clips from text descriptions. Unlike earlier video generation models, which were limited in quality and length, Sora demonstrates improved abilities in handling complex scenarios and maintaining coherence in videos up to a minute long. This model represents a significant step forward in video generation technology, combining diffusion models and transformers to process video data in new ways. However, OpenAI has expressed caution regarding the potential misuse of photorealistic video and is initially sharing the model with third-party safety testers and select video makers for feedback before considering a public release​ (MIT Technology Review)​.

These developments underscore OpenAI's commitment to advancing AI technology while also addressing the ethical considerations associated with more powerful and realistic generative models.

You

Search the web and give me the latest updates about OpenAI's Sora. When will it be available to the public?

ChatGPT

As of now, OpenAI has not announced a specific release date for Sora, its groundbreaking text-to-video AI model. Currently, Sora is in a limited release phase, accessible only to a select group of artists, filmmakers, and experts assessing its potential risks. This cautious approach reflects OpenAI's commitment to thoroughly evaluate Sora's safety and ethical implications before a broader public release​ (CCN.com)​.

Speculation suggests a potential public release could occur in 2024, but likely not before the second half of the year. This timeframe remains speculative and is subject to change based on OpenAI's assessment and preparation processes. OpenAI is focused on refining Sora's safety protocols and ensuring the technology adheres to ethical standards. Additionally, the organization must ensure it has the necessary resources and server capacity to support user demand upon the model's public release​ (Stealth Optional)​.

For the latest updates on Sora's release and the opportunity to potentially access the AI upon its public unveiling, it's recommended to follow OpenAI's official channels, such as their Twitter account​ (Stealth Optional)​.