Sciences.com: Generative Artificial Intelligence angel or demon? We talk to Jose Antonio Gamez Martin

2023-08-02 18:41:58

Artificial intelligence (AI), says our guest, José Antonio Gámez Martín, consists of making computers solve problems in such a way that, if we humans solve them, we are considered intelligent, although that solution does not have to be the same that I would find a human.

For ten years, almost everything we have heard about Artificial Intelligence is what is known as “machine learning”, that is, computer programs fed with a very high volume of data, which learn with them to solve problems intelligent. But machine learning is only one part of artificial intelligence, comments José A. Gámez, behind it are deep neural networks. Neural networks are a computational model inspired by the human brain, that is, made up of layers of units, called neurons, which process information and adjust their connections among themselves, as do the neurons in our brains. Deep neural networks have a high volume of highly interconnected neurons and are the ones that, together with machine learning, have provided innovative solutions to problems. Thus, from the automatic learning of deep neural networks, what we know as Artificial Generative Intelligence has emerged. These include models such as ChatGPT, DALL·E, BERT, etc.

We must imagine generative artificial intelligence (AI) as a development that allows machines to create original and realistic content. With ChatGPT, for example, you can chat interactively, answer questions asked, provide information, write creative or technical content, etc. DALL.E can create illustrations of objects, animals, people, landscapes, and more, all from natural language descriptions. These are just two examples of many that can produce images, music, text and even video, products that are very difficult to distinguish from those made by humans. Thus, we can say that the “machines” have snatched from the human being the monopoly of creativity that has accompanied our species throughout history.

However, as with all the inventions we have created since ancient times, generative AI also poses ethical challenges, as it can be used to create fake news, misleading content, or even spoofing. Ultimately, the perspective on whether generative AI is an “angel or devil” depends on how it is used and the values ​​and ethical standards that apply to its development and use.

Today we are talking with José Antonio Gámez Martín, Professor of Computer Systems at the UCLM School of Computer Engineering, about the universe of possibilities offered by generative AI, its fields of application, advantages, dangers, and ethical issues. I invite you to listen to him in this new chapter of Talking to Scientists.

Suggestions:

Intelligent systems and data mining. We spoke with José Antonio Gámez Martín.