Chat GPT: generative artificial intelligence within everyone's reach

An article by Giuseppe Riva, Director of the Humane Technology Lab 

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 began 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 have 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 was 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 more and more realistic content. For example, some image generation algorithms are able to generate photorealistic images of people and objects that don't actually exist. Other text generation algorithms are capable of generating text that is difficult to distinguish from text written by a human. One of the most famous generative algorithms is GPT (Generative Pre-trained Transformer), a linguistic generative AI model developed by OpenAI. The first version of GPT was unveiled in 2018, and since then there have been several subsequent versions with ever-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 the relationships between words. Thanks to this pre-training, GPT is able to generate plausible and coherent text on a wide range of topics, without the need for additional specific training. The latest version of GPT and its conversational version – ChatGPT – have sparked a great deal of interest due to its ability to generate very plausible and coherent text, which is often hardly distinguishable from that written by a human. Current applications of GPT include text generation, question answering, code generation, machine translation, and content generation.

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