OpenAI announced at its first DevDay new models and products for developers. Among them is the release of GPT-4 turbo and the possibility to develop a customised ChatGPT wizard called “GPTs”.
New releases at OpenAI
GPT-4 Turbo is an advanced version of the GPT-4 model released earlier this year. GPT-4 Turbo has updated knowledge up to April 2023 and a 128k context window, allowing it to process the equivalent of more than 300 pages of text in a single message. This model is already available for testing by paying developers, and a stable production version is expected to be released in the coming weeks.
In addition, OpenAI is committed to protecting its ChatGPT Enterprise customers and its developer platform with the introduction of “Copyright Shield”. This initiative includes the defence and payment of legal costs in the event of copyright infringement claims.
What are GPTs?
These GPTs are easy to create and can be tailored to individual or business needs, such as collecting recipes or analysing source code. They are easy to create, no coding is required and users can define their functions.
In addition, OpenAI will launch a GPT Store to host, develop, promote and evaluate these customised GPTs. Creators will be able to generate revenue through this platform. This approach not only expands OpenAI’s capabilities in the AI market, but also drives accessibility and customisation in AI technology, marking a significant step towards an ecosystem of derivative applications.
How to create a GPT in ChatGPT?
To create a GPT you just need to have a conversation with GPT Builder in ChatGPT. It will add instructions, upload files to your knowledge base and select capabilities such as web search or data analysis. It should be noted that this feature is currently available in ChatGPT Plus.
In addition, ChatGPT makes public templates created by other users available to the user, where you will also find the templates created by you.
Here’s how you can create your personalised GPT assistant step by step:
- Select “explore” and start customising your GPT by selecting the first option that appears on the screen.
- Are you clear about what you want a GPT for? We have decided to create a data professional to help us generate technical content. All we have to do is explain in words what we want and ChatGPT will create a wizard customised to our needs.
- It then suggests a GPT name and an image to identify our personalised model.
- Once we give it our ok, it starts to outline the capabilities and limitations we want to add to the GPT model. At this point it is necessary to have identified what we want for our model.
- Define the tone and the type of response that best fits our model.
- Once you save the model, you can generate the content you need with your personalised wizard adapted to the needs of your business. You can choose to publish the created model, which can be accessed through a link or leave it private so that only you can access it.
- That’s it! We have now created our personalised assistant. You can find a summary of all the features of the template in “configure”. In the configuration you can modify what you need and check that all the information generated by ChatGPT is correct.
Do you want to be updated about ChatGPT and other Generative IAs?
Subscribe to our newsletter where you can find out the latest about ChatGPT and other AI-based technologies.