Click the Card below and follow the “CVer” public account
AI/CV heavy-duty information, delivered as soon as possible
Click to enter->[Computer Vision and Transformer] Communication Group
Reply in the background of the CVer WeChat public account: ChatGPT, you can download the most complete learning materials of AIGC and start learning!
Reprinted from: Heart of the Machine | Editor: Du Wei, Da Panji
OpenAI CEO Sam Altman vowed, “We’re going to ship some great new stuff.”
In September this year, OpenAI officially announced its first developer conference “OpenAI DevDay”. OpenAI team members will gather with developers from around the world to preview new AI tools.
OpenAI CEO Sam Altman said at the time that GPT-5 or GPT-4.5 or similar large models would not be released at this developer conference. Even so, the AI tools released at the conference still aroused widespread expectations.
A few days ago, Sam Altman whetted people’s appetite again, saying that OpenAI will bring “some great new things.”
In two days, OpenAI’s first developer conference is coming. There is no airtight wall in everything, and there are still breaking news about what OpenAI is going to release, which has triggered heated discussions among netizens.
The source comes from X user CHOI, who said that OpenAI will announce a major update to ChatGPT, including a new interface and some new features: custom chatbots, connectors with Google and Microsoft, and new Subscription model.
Beyond that, OpenAI is unlikely to release a GPT4-V API.
Jim Fan, a research scientist at Nvidia who will be attending the OpenAI developer conference, said the leak, if true, would be a pivotal moment for the consumer market for AI.
OpenAI is developing into a mature UGC platform on which users can create and share any AI agent. OpenAI is also a superset of RPA, custom AI, and plugin stores. This is definitely what an AI consumer ecosystem should look like.
This also makes people look forward to what will be released at the OpenAI Developer Conference.
OpenAI’s first developer conference will release these things?
Tipster CHOI shared the full list of leaked features. According to him, OpenAI will release a Gizmo tool specifically for creating, managing and selecting custom chatbots.
Gizmo is expected to bring the following features:
-
Sandbox: Provides an environment to import, test and modify existing chatbots
-
Custom Actions: Use the OpenAPI specification to define additional functionality for your chatbot
-
Knowledge files: Add other files that the chatbot can reference
-
Tools: Provides basic tools for browsing the web, creating pictures, etc.
-
Analytics: View and analyze chatbot usage data
-
Drafts: Save and share drafts of the chatbot you create
-
Publish: Publish the completed chatbot
-
Sharing: Set up and manage chatbot sharing
-
Marketplace: Browse and share chatbots created by other users
Additionally, there is Magic Creator or Magic Maker to help you create chatbots:
-
Define chatbots through interactive interfaces
-
Identify user intent and create chatbots
-
Live test created chatbot
-
Modify chatbot behavior through iterative conversations
-
Share and deploy chatbots
Meanwhile, X user Tibor Blaho showed us more previews of the new ChatGPT prototype Gizmo V8.
Scan the QR code to join CVer Knowledge Planet, you can quickly learn the paper ideas from the latest top conferences and journalsand CV from entry to Proficient in information, as well as cutting-edge projects and applications! Highly recommended for paper submission!
He said Gizmo V8, a GPT-4 All Tools model that allows ChatGPT to browse the web, analyze data and generate images, is still in development ahead of a possible release next week. Plus faster access to DALL-E, Data Analyst and custom GPTs.
The new ChatGPT prototype looks like the ChatGPT iOS APP, but introduces a new model selector, customizable GPT and Magic Maker (possibly using GPT-4 Magic Create), and an advanced mode for creating your own GPT.
The “My GPTs” control panel displays all your GPTs and provides the option to “Create GPTs – Customize and share powerful AI assistants”.
The advanced mode of creating a new GPT allows you to specify all the details similar to the previous Gizmo prototype, including setting the profile icon, name, description, description, welcome message, etc.
New features include Prompt Starters (sample prompts for users to start conversations), Knowledge (attachment files referenced by GPT), and Actions (attachment features that GPT can use, such as web browsing, code interpreters, and other customizations). operate.)
Additionally, he shared screenshots of a “context connector” that can be used to connect other applications to ChatGPT. The screenshot below shows Google Drive and Microsoft 365. This way, a document, spreadsheet, or presentation can be attached to a chat or used as context for a conversation.
OpenAI workspace and team plans
According to the leaked information, the company will also launch new enterprise subscriptions and workspace management. Team plans are priced at $30 per month, and annual subscriptions are $25 per month.
The team plans to offer more benefits such as unlimited fast GPT-4 access, 4x longer contexts, and unlimited use of advanced data analysis models. Additionally, chat templates can be shared and OpenAI will not use the data for training purposes.
Workspace functionality separates individuals and work areas and defines roles, departments, or other metadata, and integration with external ERP and CRM systems should be possible.
Ultimately, CHOI hopes that the price of the GPT-4 API will come down, the speed of GPT-4 will increase, and the 32K version will become generally available.
As for whether these leaked information are accurate or not, we will have to wait until the developer conference to find out.
Reference link:
https://the-decoder.com/openais-massive-chatgpt-updates-leak-ahead-of-developer-conference/
https://www.reuters.com/technology/openai-make-models-cheaper-more-powerful-its-first-developer-conference-2023-11-03/
ICCV/CVPR 2023 paper and code download
Backend reply: CVPR2023, you can download the CVPR 2023 papers and code open source paper collection
Backend reply: ICCV2023, you can download the collection of ICCV 2023 papers and code open source papers
Computer Vision and Transformer exchange group established Scan the QR code below, or add WeChat: CVer444 to add CVer assistant WeChat, and then apply to join the CVer-Computer Vision or Transformer WeChat communication group. In addition, other vertical directions have been covered: target detection, image segmentation, target tracking, face detection & recognition, OCR, pose estimation, super-resolution, SLAM, medical imaging, Re-ID, GAN, NAS, depth estimation, automatic Driving, reinforcement learning, lane detection, model pruning & compression, denoising, fog removal, rain removal, style transfer, remote sensing images, behavior recognition, video understanding, image fusion, image retrieval, paper submission & communication , PyTorch, TensorFlow and Transformer, NeRF, etc. Be sure to note: Research direction + location + school/company + nickname (such as target detection or Transformer + Shanghai + hand in + Kaka). Note according to the format to get passed and invited to the group faster
▲Scan the QR code or add WeChat ID: CVer444 to join the communication group CVer Computer Vision (Knowledge Planet) is here! If you want to know about the latest, fastest and best CV/DL/AI paper express delivery, high-quality practical projects, AI industry cutting-edge, and learning tutorials from entry to mastery, please scan the QR code below and join CVer Computer Vision (Knowledge Planet). Nearly ten thousand people have been gathered! ▲Scan the QR code to join Planet Learning
▲Click on the card above to follow the CVer official account
It’s not easy to organize, please like and watch