What is Generative AI?

What is Generative AI?

• Traditional or conventional AI typically focuses on specific tasks and is designed to solve specific problems. It operates based on predefined rules and algorithms, often with a fixed set of inputs and outputs. AI systems are trained to follow specific instructions and make decisions based on predetermined logic.

• Generative AI is focused on creating new and original content. It is capable of generating outputs that are not explicitly programmed or predefined. Generative AI models are trained on very large datasets to learn patterns and structures, enabling them to generate new and creative content that resembles the training data. They can produce text, images, music, speech, video, and other forms of content based on the patterns and examples they have learned.

How do LLMs work?

• Large Language Models (LLMs) mirror characteristics of natural language and competent writing on a wide range of subjects in response to a user request. 

• LLMs have been trained on vast quantities of text from nearly everything ever written and posted on the internet. Using this data, an LLM generates original text responses using a statistical algorithm that predicts the next word. The more data it is trained on, the better and more natural the language responses are. 

• While LLMs like ChatGPT do not have a sense of "facts" or "truth," advanced models can access the internet to provide real-time information and verify current events.

What tools are supported at NYU?

Using your own personal subscription to an LLM can be convenient, but it introduces concerns about privacy, security, and access.

NYU and Stern offer Google Gemini for all students and faculty. 

A key advantage of using this Stern-supported platform is that its protections are governed by the same privacy and security agreements that apply to other licensed Google services at NYU. It also ensures equal access to the service among your students. We recommend, where possible, using Google Gemini with your NYU Stern account when you incorporate LLMs into your assignments or classwork.

NYU and Stern’s access also includes NotebookLM. NotebookLM allows users to synthesize information from their own documents. By grounding itself in sources you upload, it can answer questions, help students locate content in their sources, and even structure study guides. Many faculty find NotebookLM is better aligned with academic use than other AI tools.

Learn more about NYU’s supported AI services here.