Generative AI (GAI) is a type of artificial intelligence that can generate text, images, videos, and other content in response to a user prompt, based on its training data.
This guide focuses on text-based AI generators. For information on image and video AI generators, see the Machines and Society guide.
Large language model (LLMs) are the foundation of GAI. LLMs are trained on vast amounts of text to understand existing content and generate original content. They function as chatbots, responding to user prompts by processing natural language in a conversational, human-like way. They can perform a variety of language-based tasks, like generating, summarizing, and translating text.
ChatGPT: (OpenAI) Has free and fee-based options
Copilot: (Microsoft) Free
Gemini: (Google) Free to anyone with a Google account
Llama2: (Meta) Free for download
When asked what it is capable of doing, this is one way ChatGPT has answered:
(ChatGPT, personal communication, March 14, 2023)
What is behind ChatGPT's mystique? This is one way that ChatGPT has responded to the question of how it works:
ChatGPT is a language model that utilizes a range of concepts, architectures, processes, and data inputs to perform its tasks. Here are some of the key elements that enable ChatGPT to work:
Large language model: ChatGPT is a large language model that is designed to process and generate large amounts of natural language text. It is trained on massive amounts of text data, including books, articles, websites, the Common Crawl, Wikipedia, and other sources. These datasets provide the foundational knowledge that ChatGPT draws on when processing input text and generating responses.
Artificial Neural Networks (ANNs): ANNs are mathematical models that simulate the way that the human brain processes information. They are used extensively in ChatGPT to process input text and generate responses.
Natural Language Processing (NLP): This is the use of computational techniques to analyze and understand human language. ChatGPT uses NLP to process input text and generate responses.
Transformer architecture: The transformer architecture is a type of neural network that is used in GPT models. It enables ChatGPT to process input text and generate responses in a way that is both efficient and effective.
Generative Pre-trained Transformer (GPT): ChatGPT is based on GPT, a type of deep learning model specifically designed for natural language processing. GPT is trained on vast amounts of text data to learn the patterns and structures of language, enabling it to generate human-like responses to input prompts.
Transfer learning: ChatGPT is trained on a vast corpus of text data that covers a wide range of topics and styles. Transfer learning enables ChatGPT to transfer knowledge learned in one domain to another related domain, allowing it to generate responses to a diverse range of input prompts.
Reinforcement learning: ChatGPT uses reinforcement learning to improve its performance over time. This involves training it on a dataset of prompts and human-generated responses, and then adjusting its behavior based on feedback from humans.
Attention mechanism: The attention mechanism is a component of the transformer architecture that allows ChatGPT to focus on specific parts of the input text when generating a response. This helps it to generate more accurate and relevant responses.
Overall, ChatGPT utilizes a complex combination of technologies and processes to analyze and understand human language, enabling it to generate responses that are similar to those produced by humans.
(ChatGPT, personal communication, March 14, 2023)
User question: Can you provide a diverse list of the best books of the 20th century?
ChatGPT answer: The 20th century produced an incredible amount of great literature across a wide range of genres and styles. Here are some of the best books of the 20th century, in no particular order, that represent a diverse range of voices and perspectives:
This is by no means an exhaustive list, but it includes some of the most influential and widely-read books of the 20th century.
(ChatGPT, personal communication, March 14, 2023)
User question: What is the most efficient cardio-vascular exercise?
ChatGPT answer: There is no single "most efficient" cardio-vascular exercise as different exercises offer different benefits and are suited for different individuals and fitness goals. However, some cardio exercises are generally considered more effective than others for improving cardiovascular fitness and burning calories.
Here are a few examples of popular cardio exercises that are generally considered effective:
Ultimately, the most efficient cardio exercise for you will depend on your individual fitness level, goals, and preferences. It's important to choose an exercise that you enjoy and can stick to in the long term, as consistency is key to seeing results.
(ChatGPT, personal communication, March 13, 2023)
User question: What should I make for dinner tonight?
ChatGPT answer: I'm happy to help you come up with some ideas for dinner tonight! Here are a few suggestions based on different dietary preferences and ingredients you might have on hand:
(ChatGPT, personal communication, March 13, 2023)