Chat GPT: generative artificial intelligence within everyone's reach

Article by Giuseppe Riva, Humane Technology Lab Director

Generative AI is a subcategory of artificial intelligence that focuses on using algorithms to generate content, such as images, text or sounds, instead of simply recognizing or classifying it. The history of generative AI begins in the 1950s, when the first automatic text generation algorithms were created. Initially, these algorithms were limited and produced only simple, repetitive sentences. Over the years, developers continued to work on these algorithms to improve their ability to generate more realistic and plausible text.

In the 1990s, with the advent of more powerful computers and machine learning, it became possible to use generative AI to generate images and sounds as well. Over the past few years, generative AI has become increasingly sophisticated and capable of generating increasingly realistic content. For example, some image generation algorithms are able to generate photorealistic images of people and objects that do not exist in reality. Other text generation algorithms are able to generate text that is difficult to distinguish from text written by a human being. One of the most famous generative algorithms is GPT (Generative Pre-trained Transformer), a linguistic generative artificial intelligence model developed by OpenAI. The first version of GPT was presented in 2018, and since then, there have been several subsequent versions with increasing capabilities. 

The model was trained on a large corpus of text, using an unsupervised learning algorithm called "pre-training," so that it could gain a general understanding of language and word relationships. Because of this pre-training, GPT is able to generate plausible and coherent text on a wide range of topics, without the need for further specific training. The latest version of GPT and its conversational version - ChatGPT - have attracted a great deal of interest because of its ability to generate very plausible and coherent text that is often difficult to distinguish from text written by a human being. Current applications of GPT include text generation, question answering, code generation, machine translation, and content generation.

To highlight ChatGPT's creative capabilities, we asked it questions that it answered without any problems.

If we were to explain to a non-expert user what ChatGPT is, what could we tell him?

ChatGPT is a large language model trained by OpenAI. It is used to generate text autonomously, such as by answering questions or generating writing. This means it can understand and develop human language similarly as a human would.

But can we talk to CHATGPT about anything?

ChatGPT has been trained on a wide range of texts; therefore, it has general knowledge of many topics. However, as with any machine learning model, its understanding and ability to answer a specific question will depend on the amount and quality of data it has been trained on. So while it can be asked many things, it cannot always give an accurate or complete answer.

For example, if we ask you to come up with a poem about student problems, can you write it?

Yes, I can try to write a poem that talks about students' problems. Here's an example:

Anxiety grows with each class,
studying becomes a pressure.
The mind is full of doubt and fear,
the future seems uncertain.

Time flies by quickly,
time is too short for us,
homework seems a cross,
sleep is a luxury we cannot afford.

Exams make us tremble,
the judgment of others makes us suffer,
but despite everything, we want to arrive,
to the goal, we set for ourselves.

Students are like delicate flowers,
growing up in a world full of stress,
but if we give ourselves the strength to endure,
we can become what we want to be.

Of course, this poem is just an example and may not be the one you were looking for, but I hope you enjoyed it.