Artificial Intelligence

What is Generative AI? 

Generative AI refers to artificial intelligence technology that can produce various types of content, such as text, imagery, audio, and synthetic data based on the data it was trained on. It involves deep-learning models that can generate high-quality content like text and images. Generative AI models, like ChatGPT and DALL-E, have the ability to create new and original content autonomously, making them valuable for creative fields and novel problem-solving. These models are trained to generate outputs that resemble the data they were trained on, allowing them to produce diverse forms of content from music and art to virtual worlds. Generative AI is powered by large AI models that can perform various tasks like summarization, Q&A, and classification with minimal training required. It works by learning patterns and relationships in human-created content datasets to generate new content that aligns with the learned patterns.

What are large language models?

Large language models (LLMs) are advanced deep learning algorithms capable of understanding, summarizing, translating, predicting, and generating content using extensive datasets. These models, like OpenAI's GPT series and Google's BERT, are trained on vast amounts of text data to develop a foundational understanding of human language. LLMs leverage transformer architectures, which consist of encoder and decoder components, to process data by tokenizing input and identifying relationships between tokens. They utilize self-attention mechanisms to learn quickly and generate predictions based on the context of a sentence. LLMs are composed of various neural network layers like recurrent layers, feedforward layers, embedding layers, and attention layers that work together to process input text and produce output content. These models have billions of parameters that act as a knowledge bank, enabling them to capture intricate language patterns and generate coherent responses across various tasks such as translation, chatbots, AI assistants, creative writing, and code generation. LLMs are revolutionizing applications in fields like chatbots, virtual assistants, content generation, research assistance, and language translation.

These definitions were generated using Perplexity AI.

PerplexityAI. (2023). Perplexity [Large language model]. https://www.perplexity.ai

• Bias - AI tools can inherit bias from the systems on which they are trained.
• Hallucinations - Because the content is generative, AI tools may create false content that appears genuine.
• Privacy - Many of these tools capture user data or input, which causes privacy concerns.
• Copyright - There are several aspects of copyright to be considered with the use of generative AI. Concerns have been expressed over the copyright of the content on which LLMs are trained. Also, it is good practice to cite the use of generative AI.