Artificial Intelligence

Artificial Intelligence: Principles, Prediction, Security

Instructor: Mgr. Václav Peca

Location: Gymnázium Brno, třída Kapitána Jaroše, Public Institution

Time and program of individual meetings will be specified in September.

Through practical activities, we will learn how neural networks and AI systems based on them work, evaluate their security, and try to predict the future development of AI technology based on data extrapolation and understanding of basic principles.

We will train our own simple neural networks and understand how to go from basic mathematical principles (such as gradient descent) and simple linear regression, through a classic network for recognizing numbers from the MNIST dataset, to today's large language models like GPT.

We will try working with various AI tools and learn to evaluate how safely and effectively they function (including automatic evaluations using APIs).

Based on the extrapolation and aggregation of available data and knowledge (ML benchmarks, computational performance development, expert predictions, theoretical studies, ML experiments), we will try to predict the future development and implications of AI technology, including transformative scenarios (AGI - artificial general intelligence, HLMI - human-level machine intelligence).

The program includes: