OpenAI, the organization renowned for its revolutionary ChatGPT model, is celebrating the season with its unique event dubbed "Shipmas," which commenced on December 6 and is set to continue for twelve days. During this event, OpenAI reveals remarkable new products, features, and enhancements to its AI technology aimed at improving user experience and engagement.
Day 1 of the event unveiled the OpenAI o1 model for ChatGPT, alongside the introduction of ChatGPT Pro, a subscription service aimed at businesses priced at $200 per month. This revolutionary model is outlined by OpenAI as enhancing user experience by breaking down inquiries and tasks, allowing it to address each component before delivering thorough responses. Remarkably, it has been reported to solve 83% of questions from the prestigious International Mathematics Olympiad, with increased accuracy and faster response times compared to its predecessor. The o1 model also introduces image processing capabilities, enabling users to upload photos for analysis and guide creation.
On Day 2, the spotlight was on developers with the launch of AI Reinforcement Fine-Tuning, which enhances the performance of AI models by refining their reasoning capabilities for improved responses. This enhancement serves as a valuable tool for developers who seek to customize OpenAI’s algorithms for more complex tasks, particularly beneficial for sectors like law and science.
Moving to Day 3, OpenAI launched Sora, its much-anticipated AI image generation tool, initially teased over a year ago. Now accessible to ChatGPT Plus and Pro members, Sora employs the upgraded Sora Turbo model which offers faster and high-quality video production. Users can create videos from text prompts, animate still images, or remix videos according to their requirements, with every subscriber receiving 1,000 monthly credits to create videos at varying lengths and resolutions.
Day 4 saw the introduction of Canvas, enhancing collaboration with code and text. This innovative feature allows users to receive and share responses from the AI on a shared interface, facilitating joint efforts on writing and coding tasks. Developers can benefit from Canvas by receiving suggestions, improvements, and debugging assistance directly from ChatGPT.
On Day 5, OpenAI integrated ChatGPT with Apple Intelligence as part of iOS 18.2, aiming to enrich Siri's capabilities. While this announcement was overshadowed by a temporary outage affecting multiple services including ChatGPT and Sora, the functionality was restored shortly thereafter, returning to normal operations.
Midway through the event, on Day 6, OpenAI introduced Advanced Voice Mode for ChatGPT, featuring visual capabilities. This enhancement allows users to share their screen and stream video directly to the AI, mirroring the experience of consulting with a friend. Through this mode, users can interact with ChatGPT hands-free, creating opportunities for more natural dialogues.
Day 7 was all about organization, as OpenAI launched Projects—a new feature enabling users to categorize their interactions with ChatGPT. This function serves as a virtual filing system for users, allowing them to create organized projects where conversations and files can coexist. When starting a project, users assign it names and colors, making management easier.
On Day 8, OpenAI expanded its capabilities by introducing the ChatGPT Search function. This feature integrates advanced voice mode functionalities with precise web searching, allowing the AI to fetch updated information or draw from its existing database when answering inquiries.
Day 9 highlighted OpenAI o1's latest announcement to developers, with support for API access, which enhances integration possibilities with external data and APIs. This feature paves the way for real-use applications across various fields, particularly scientific domains.
On Day 10, OpenAI introduced call and WhatsApp messaging support for ChatGPT, allowing users to interact via voice or text without needing to log in or set up accounts. This experimental service is available to users aged 13 and above, providing 15 minutes of free call time monthly and daily messaging, broadening user accessibility.
The penultimate day of Shipmas, on Day 11, saw OpenAI announcing enhanced support for ChatGPT with compatibility across multiple coding and note-taking applications on macOS. This beta feature optimizes workflow by embedding ChatGPT directly within users' desktop environments, allowing seamless engagement with applications like Apple Notes and Xcode.
Finally, Day 12 concluded the event with the introduction of the new reasoning models, o3 and o3 Mini. While there are no immediate plans for public release, OpenAI CEO Sam Altman announced expectations for the o3 Mini model to launch by January 2025, marking another advancement for the organization.
With such extensive developments made during the Shipmas event, OpenAI is setting the stage for future advancements, promising users enhanced capabilities and functionality as it continues to refine its groundbreaking AI technologies.
