At OpenAI DevDay, the company’s first-ever developer conference held on November 6, OpenAI CEO Sam Altman announced that a hundred million people are using ChatGPT on a weekly basis and there are over two million developers, including over 92 percent of Fortune 500 companies. The company also made the following major announcements at the conference:
GPTs: Custom versions of ChatGPT
OpenAI will allow users to create custom versions of ChatGPT called GPTs that are tailored for specific use cases. “GPTs are a new way for anyone to create a tailored version of ChatGPT to be more helpful in their daily life, at specific tasks, at work, or at home—and then share that creation with others,” the company explained. Examples include chatbots specific for trip planning, marathon training, or creative writing.
Here's a little GPT (the name for the new agent-like-thing released by Open AI) that I threw together in less than a minute.
It looks up the latest trends for a product category on the web and then creates prototype images for it. Takes less than 90 seconds end-to-end pic.twitter.com/pbflSJn3Gh
— Ethan Mollick (@emollick) November 6, 2023
OpenAI says that users can create GPTs without any coding skills. “Creating one is as easy as starting a conversation, giving it instructions and extra knowledge, and picking what it can do, like searching the web, making images or analyzing data.”
GPTs can also be integrated with external data or interact with the real world. “Connect GPTs to databases, plug them into emails, or make them your shopping assistant. For example, you could integrate a travel listings database, connect a user’s email inbox, or facilitate e-commerce orders,” the company said.
Later this month, users will also be able to upload GPTs they create to the upcoming GPT Store. “In the coming months, you’ll also be able to earn money based on how many people are using your GPT,” OpenAI said.
GPTs are available for ChatGPT Plus and Enterprise users.
New GPT-4 Turbo and updated GPT 3.5 Turbo
OpenAI launched the next generation of its GPT-3 model, the GPT-4 Turbo. “GPT-4 Turbo is more capable and has knowledge of world events up to April 2023. It has a 128k context window so it can fit the equivalent of more than 300 pages of text in a single prompt,” the company explained.
GPT-4 Turbo is available in preview for all paying developers.
OpenAI also released an updated version of GPT-3.5 Turbo. “The new 3.5 Turbo supports improved instruction following, JSON mode, and parallel function calling,” the company said.
Lower pricing and higher rate limits
OpenAI is decreasing prices across the platform:
- GPT-4 Turbo input tokens are 3x cheaper than GPT-4 and output tokens are 2x cheaper.
- GPT-3.5 Turbo 16K model input tokens are 3x cheaper than the previous 16K model and output tokens are 2x cheaper at $0.002.
- GPT-3.5 Turbo 4K model are 33% cheaper on input tokens at $0.001.
- Fine-tuned GPT-3.5 Turbo 4K and 16K model input tokens are reduced by 4x at $0.003 and output tokens are 2.7x cheaper at $0.006.
Additionally, OpenAI is doubling the tokens per minute limit for all its paying GPT-4 customers.
Immunity from copyright lawsuits
Joining companies like Microsoft, Google, and Adobe, OpenAI said that it will defend its customers who face copyright lawsuits for using OpenAI’s products. Under the Copyright Shield program, “we will now step in and defend our customers, and pay the costs incurred, if you face legal claims around copyright infringement. This applies to generally available features of ChatGPT Enterprise and our developer platform,” OpenAI said.
- Information up to April 2023: ChatGPT Plus now includes fresh information up to April 2023.
- Assistant API: OpenAI launched Assistant API (Application Programming Interface) that will allow developers to build agent-like experiences within their applications. “An assistant is a purpose-built AI that has specific instructions, leverages extra knowledge, and can call models and tools to perform tasks,” the company explained. “This API is designed for flexibility; use cases range from a natural language-based data analysis app, a coding assistant, an AI-powered vacation planner, a voice-controlled DJ, a smart visual canvas—the list goes on,” it added.
- GPT-4 Turbo will accept image inputs: “GPT-4 Turbo can accept images as inputs in the Chat Completions API, enabling use cases such as generating captions, analyzing real world images in detail, and reading documents with figures,” OpenAI informed.
- DALL-E integration with apps: “Developers can integrate DALL·E 3, which we recently launched to ChatGPT Plus and Enterprise users, directly into their apps and products through our Images API.”
- Text-to-speech generator: Developers can use the new text-to-speech API to generate human-quality speech from text.
- Fine-tuning and custom models: Similar to GPT 3.5, users will be able to fine-tune GPT-4. However, there will be an experimental access program for GPT-4 fine-tuning because “preliminary results indicate that GPT-4 fine-tuning requires more work to achieve meaningful improvements over the base model compared to the substantial gains realized with GPT-3.5 fine-tuning,” the company said. For organizations that need even more customization than fine-tuning OpenAI is launching a Custom Models program. “This will be a very limited (and expensive) program to start,” the company said.
- Simpler model picker: “We’ve also heard your feedback about how the model picker is a pain. Starting today, no more hopping between models; everything you need is in one place. You can access DALL·E, browsing, and data analysis all without switching. You can also attach files to let ChatGPT search PDFs and other document types,” OpenAI said.
STAY ON TOP OF TECH NEWS: Our daily newsletter with the top story of the day from MediaNama, delivered to your inbox before 9 AM. Click here to sign up today!
- UK’s Bletchley Declaration Calls For Global Cooperation In Identifying AI Risks, Ensuring Safety
- 223: The Number Of Times AI Was Mentioned In Google, Microsoft, And Meta Earnings Calls
- Google Will Protect Its Generative AI Users From Copyright Lawsuits
- AI, A Derivative Of The Surveillance Business Model: In Conversation With Signal President Meredith Whittaker