Summary
In this chapter, we looked at ANNs and how they evolved from simple concepts to powerful AI models. We also had a run-through of the most popular architectures in use today, starting from the humble perceptron and finishing with the powerful transformer model. Finally, we skimmed through the complexities and limitations of ANNs to get an overview of the open challenges that still exist – hence why we need to seek new, more powerful alternatives. Because of this, in our next chapter, we will look at explainable AI and its various approaches that help to build trust in AI systems.
