OpenAI has released a set of new features with artificial intelligence advancements for developers and ChatGPT users. These users can now create custom GPTs with tons of improvements to their models.
Quick Take:
- OpenAI comes up with the best GPT version to date, GPT 4 Turbo.
- The supercharged tool stores information until April 2023 and supports up to 128K tokens.
- GPT 4 Turbo comes with a more detail-oriented approach towards user prompts.
- OpenAI unveils updates to GPT 3.5 and launches Custom GPTs and Assistants API.
At its annual DevDay conference, OpenAI unveiled several key changes to ChatGPT and launched other products. The updates include wholesale price reductions for developers and users. The company announced the launch of a creator tool for chatbots, called GPTs.
Most importantly, OpenAI came up with a new language model called Turbo. The platform consistently comes up with new and advanced updates for ChatGPT. Earlier this year, it rolled out GPT 3.5 and GPT 4. This time around, the platform has launched the GPT 4 Turbo Version.
Key Aspects of GPT 4 Turbo Version
The launch of GPT 4 Turbo was the showstopper of OpenAI’s conference. It is a supercharged version of the existing AI model. Moreover, it offers increased efficiency and a more in-depth understanding of current world events and scenarios.
Upon the announcement, the CEO of OpenAI, Sam Altman said,
“We are just as annoyed as all of you, probably more, that GPT-4’s knowledge about the world ended in 2021.”
Therefore, with knowledge of events until April 2023, GPT 4 Turbo can give more relatable and accurate responses to user prompts. However, the use of information by ChatGPT remains a point of concern for authors and publishers. They believe that their content is used by the platform without their consent.
GPT 4 Turbo – 128K Tokens and Reduced Prices
Another key feature of GPT 4 Turbo is that a user can now input more information at once, as the tool supports up to 128,000 tokens of context. Altman compared it to being around the number of words in 300 book pages. This will help users analyze and get summaries of long texts and transcripts.
As per details from OpenAI, GPT 4 Turbo will also be a better listener than all previous GPT models. It will pay more attention to details requested by the user in their prompts. It will be more useful for users who are trying to create formats and generate codes with the chatbot.
For developers, GPT 4 Turbo will come at a much reduced price. Its price is one cent for a thousand prompt tokens and three cents for a thousand completion tokens. Therefore, developers will be able to use the tool with less information and get their desired answers.
In another update, OpenAI has scrapped the drop-down menu to select chatbot tools. Be it Dall-E 3 or Bing, GPT 4 Turbo will select the right tools to answer the user prompt. Users do not need to opt for any tools beforehand.
OpenAI Lays Foundation of Custom GPTs and Assistants API
The tech firm has also come up with a new product, called GPTs. It is a customized version of ChatGPT that users can create for any particular purpose. OpenAI aims to facilitate users creating these GPTs on their own and sharing their creations with others.
Many OpenAI employees created their own chatbot agents at the DevDay conference. Most importantly, it does not require any coding skills or knowledge. Anyone can create these tools to help with data analysis or web crawling for information. The product is available to ChatGPT Plus and Enterprise users.
Furthermore, OpenAI has given birth to the Assistants API. The new tool will help in the creation of AI-driven applications. Users can create applications that can handle complex tasks efficiently. In conclusion, the firm has taken a step ahead by exploring and introducing the usability of AI for consumers.
