From the course: OpenAI API for Python Developers

Generative AI: Genesis and evolution

From the course: OpenAI API for Python Developers

Generative AI: Genesis and evolution

“

- [Instructor] AI is a very broad field with the ambitious goal to create intelligence machines. Over the years, the applications of AI have split between multiple subsets, including machine learning and deep learning. Machine learning is a subset of AI to give computers the capacity to learn through training data to make projections, and it has been around since the early days of computing in an effort to leverage technologies around artificial intelligence. And this is in 1966 that the first ever AI chatbot was created, Eliza, which was capable of simulating the same interactions and ask the same questions of a psychotherapist. So progressively, we evolved from supervised to unsupervised machine learning with deep learning. And deep learning is another subset of AI and machine learning combined that gives machine the ability to self-improve and to learn from their mistakes with a minimum of supervision. So instead of giving computers specific instructions, computers can figure out what actions to take next to achieve the end result. In that sense, AI is making machines capable of taking the commands of a humans, like in the automotive industry to build self-driving cars or the autopilot systems in airplanes. Years later and today, AI is progressively entering our day-to-day lives and work with more AI-driven tools and services. And without even knowing it, our online experience has been powered by machine learning with personalized content and tailored recommendations when watching streaming content, and when we use translation tools. Also with customer support, these chatbots can provide instant assistance and also escalate issues to human agents whenever it is necessary. So customer service chatbot use natural language processing, NLP, and machine learning to understand and respond to customer queries. And AI is also integrating our home with personal digital assistance like Apple Siri, Alexa, or Google Home that are capable of performing actions on voice commands. And now with ChatGPT, that literally took the internet by storm since its release in November, 2022. And ChatGPT is a chatbot, but not just any chatbot. It is what we call a large language models, optimized for dialogue that understands natural language and that can perform various text processing tasks. And it is powered by the GPT models that are types of neural network based models inspired by the neurons of a human brain. Much like the human brains, large language models are capable to understand, learn, and to self-improve, to generate new content based on what they've learned. And so generative AI is another subset of AI machine learning and deep learning altogether, giving applications like ChatGPT, the ability to generate new content, like text, images, music, or other types of data. And also with the combination of NLP, natural language processing, again, another subset of AI. And there has been one domain where generative AI has been disruptive; in web developments. So generative AI offers new possibilities in the way that websites and applications are developed. So we can use generative AI to automate tasks like content creation, generate artistic images, create chatbots and virtual assistance powered by AI to answer user queries, and help troubleshoot technical issues. So we'll see together a nice example in this course. So this course explores several examples of AI integration with Python from basic to advanced recreate powerful AI agents that we set with specific behaviors. We also connect to tools like function calling in order to expand the many capabilities of large language models. We also experiment with the integration of other generative models developed by OpenAI, like moderation, in order to filter designed contents. Then we're going to see an example with image generation, with DALL-E, also with Whisper, Whisper audio to convert audio to text. And finally, we'll see an example with embeddings to allow custom knowledge retrieval. In this course, we're not building the next self-driving car, but we're sure going to have a deep dive into the integration of generative AI with practical and fun examples.

Contents