OpenAI recently concluded its “12 Days of OpenAI” event, unveiling a series of significant advancements and features from December 5 to December 20, 2024. Here’s a comprehensive overview of the announcements made during this period:
Day 1: Release of o1 Model and Introduction of ChatGPT Pro
OpenAI launched the full version of its o1 reasoning model, designed to enhance complex problem-solving capabilities through chain-of-thought processing. This model is particularly adept at tasks requiring advanced reasoning, such as mathematics and coding. Alongside this, OpenAI introduced ChatGPT Pro, a premium subscription priced at $200 per month, offering unlimited access to the o1 model and additional features like Advanced Voice Mode.
Day 2: Expansion of Reinforcement Fine-Tuning Research Program
The company expanded its Reinforcement Fine-Tuning Research program, enabling developers to train AI models to excel at specific, complex, domain-specific tasks. This initiative aims to create more adaptable AI systems capable of handling diverse real-world challenges.
Day 3: Launch of Sora Video Generation Model
OpenAI unveiled Sora, a text-to-video generation model that allows users to create realistic videos from textual descriptions. Available to ChatGPT Plus and Pro subscribers, Sora enables the creation of 20-second, 1080p video clips, marking OpenAI’s entry into AI-driven video content creation.
Day 4: Enhancement of Canvas Feature
The Canvas feature was integrated directly into the GPT-4o model, making it accessible to all users, including those on the free tier. Canvas provides a collaborative space for editing code and text, with new capabilities such as running Python code within the interface and constructing custom GPTs.
Day 5: Integration with Apple Intelligence
OpenAI announced a collaboration with Apple to integrate ChatGPT into Apple Intelligence, specifically Siri. With the release of iOS 18.2, users can invoke ChatGPT directly through iOS, enhancing the AI capabilities of Apple’s virtual assistant.
Day 6: Advanced Voice Mode Gains Visual Capabilities
Advanced Voice Mode (AVM) was upgraded to include visual input, allowing the AI to analyze surroundings through a device’s camera or screen sharing. This enables users to receive contextual assistance based on visual information. Additionally, a seasonal “Santa” voice was introduced for AVM, adding a festive touch to user interactions.
Day 7: Introduction of Projects for ChatGPT
OpenAI introduced “Projects,” a smart folder system designed to help users organize their chat histories and uploaded documents by subject. This feature aims to enhance user experience by providing better organization and accessibility of information.
Day 8: ChatGPT Search Made Available to All Users
The ChatGPT Search function, which allows the AI to search the internet for real-time information, was made available to all logged-in users, regardless of subscription tier. This feature enhances the AI’s ability to provide up-to-date and accurate responses by accessing current web data.
Day 9: Full o1 Model Accessible via API
OpenAI granted select developers access to the full version of the o1 reasoning model through its API. This move allows for the integration of advanced reasoning capabilities into third-party applications, fostering innovation and development in AI applications.
Day 10: Launch of 1-800-ChatGPT
To reach users without internet access, OpenAI introduced a telephone service, 1-800-ChatGPT (1-800-242-8478), allowing individuals to interact with the AI’s Advanced Voice Mode via phone for up to 15 minutes for free. This initiative aims to make AI assistance more accessible to a broader audience.
Day 11: Expanded Integration with Coding and Text Applications
OpenAI expanded ChatGPT’s integration with various coding applications and integrated development environments (IDEs), as well as conventional text programs like Apple Notes, Notion, and Quip. This enhancement allows users to receive AI assistance directly within their preferred applications, streamlining workflows and improving productivity.
Day 12: Preview of o3 and o3-mini Reasoning Models
On the final day, OpenAI CEO Sam Altman announced the upcoming o3 and o3-mini reasoning models. These next-generation models are designed to offer superior performance on complex benchmarks in math, science, and coding, surpassing the capabilities of previous models. The o3-mini model is available as a preview for researchers to conduct safety testing and red-teaming trials.
Check out the full announcement playlist here:
Also learn more here: https://openai.com/12-days/
