There's little we can do, either you join or you're left out. But not in a social media way (FOMO), but something more forceful.
I don't think it's time to go crazy trying to implement ChatGPT in your company (beyond the basic tasks it can help you with), but it is time to be very attentive and closely follow its evolution.
Proof of this are the new features
telegram usa number about ChatGPT that were announced at the last OpenAI developers day (the tip came to me through my friend Fernando Polo's newsletter. You can subscribe here ).
There is a lot of information on the OpenAI Keynote , but I am going to focus on the part that affects us the most from a digital marketing point of view: virtual assistants based on GPT-4 Turbo ( at the end of the article I leave you a 3-minute summary video ).
I think it will revolutionize the way we interact with our customers.
What is Retrieval Augmented Generation (RAG) and what is it used for?
The big new feature they have announced is called Retrieval Augmented Generation (RAG). It will be an incredible improvement for current chatbots. By joining with GPT-4 Turbo, the chatbot can use external information to give more precise and complete answers.
This chatbot could access the company's database to collect specific information about users or products. Can you imagine how this would change the customer experience?
I imagine them as the chatbots we have always imagined. Helpful customer service that only uses a human agent for things where they really add value: making a sale, managing a complaint, etc. etc.
In addition, to create the new chatbots they have created a user-friendly interface to connect to the OpenAI API.
And the best thing of all is that now, without needing to be a programming genius, we can use these technologies thanks to the user-friendly interface of the OpenAI API.
ChatGPT-4 Turbo Features
Retrieval : This feature allows our assistants to access external information without the need for complex configurations. This means that they can leverage the company's own data or relevant information from other sources to provide better answers.
Code interpreter : Attendees will be able to write and run Python code to create graphs and handle different types of data formats. This opens up a world of possibilities for solving complex math problems and much more.
Function Calling: This feature will allow our assistants to invoke specific functions and use the responses in their interactions. Is this the closest thing to reasoning?
I believe this will change the way companies interact with their customers and improve the digital customer experience.
It is true that we will see a wave of virtual assistants, and not all of them will work well at first or be really useful, but it is an almost obligatory first step for this technology to advance.