Artificial Intelligence: reclaim the power

Artificial Intelligence: reclaim the power

Artificial Intelligence Ethical issues Challenges and best practices

Thomas Belhalfaoui

Thomas has been working since 2013 as a Machine Learning Engineer, at the crossroads of Computer Science and Applied Mathematics. He designs and develops Artificial Intelligence algorithms for companies and public services. He also teaches these subjects and has a particular interest in the societal issues surrounding these technologies.

 

Since the recent advent of generative Artificial Intelligence systems such as ChatGPT, all we need is a click to get the answer to a question or to write a full text on any subject. But is AI actually telling the truth? How do they produce texts? How reliable are the results? Overconfidence in these technologies may lead us to produce biased and discriminatory content, and therefore to take prejudicial decisions. Is it possible to prevent this?

AI is far from being just ChatGPT: we've been surrounded by Artificial Intelligence for two decades now. Just think of the algorithms that decide what content we see every day on social networks or in search engine results. What exactly are these algorithms and how do they affect us?

In this talk, we will attempt to demystify AI, to raise awareness about its risks and learn how to make the most of it:

  • Know what an Artificial Intelligence algorithm is and understand its general operating principles - lift the lid on the "black box".

  • Become aware of the ethical issues surrounding AI systems, and their impact on our daily lives and on society.

  • Use concrete examples to analyze AI results with a critical eye and highlight its possible biases.

  • Get acquainted with the practical tools you need to make the best use of an AI such as ChatGPT, while remaining aware of its limitations. Learn, as a user, to put yourself back at the center of the process and reclaim the power over the algorithms.