A Simplr Glossary to Generative AI
Struggling to keep up with the rapidly changing landscape and terminology of generative AI? Well, you came to the right place. Welcome to our GenAI glossary – a go-to resource for unraveling the intricate world of generative AI.
In this glossary, we aim to demystify the terminology that often surrounds GenAI. Whether you’re a CX leader or industry professional, a seasoned tech enthusiast, or simply a curious learner, this guide will equip you with the essential vocabulary to navigate through the complexities of generative AI with confidence. Let’s dive in!
- Artificial Intelligence (AI): the simulation of human intelligence by machines. Computer programs that leverage copious amounts of data and compute are trained to perform tasks such as decision-making and problem-solving with minimal human intervention. AI algorithms are typically rule-based and built through iterative processing to recognize patterns and make predictions. The evolution in technology such as cloud, compute, and Big Data have been instrumental in making AI faster, cheaper, and more accessible.
- Generative AI (GenAI): a broad term that can be used for any AI system whose primary function is to generate content. This is in contrast to AI systems that perform other functions, such as classifying data (e.g., assigning labels to images), grouping data (e.g., identifying customer segments with similar purchasing behavior), or choosing actions (e.g., steering an autonomous vehicle).
- Large Language Models (LLMs): a type of AI system that works with language. In the same way that an aeronautical engineer might use software to model an airplane wing, a researcher creating an LLM aims to model language, i.e., to create a simplified—but useful—digital representation. The “large” part of the term describes the trend towards training language models with more parameters. A key finding of the past several years of language model research has been that using more data and computational power to train models with more parameters consistently results in better performance. Accordingly, cutting-edge language models trained today might have thousands or even millions of times as many parameters as language models trained ten years ago, hence the description “large.”
- Machine Learning (ML): a branch of artificial intelligence (AI) and computer science which focuses on the use of data and algorithms to imitate the way that humans learn, gradually improving its accuracy. Machine learning starts with data — numbers, photos, or text, like bank transactions, pictures of people or even bakery items, repair records, time series data from sensors, or sales reports. The data is gathered and prepared to be used as training data, or the information the machine learning model will be trained on. The more data, the better the program.From there, programmers choose a machine learning model to use, supply the data, and let the computer model train itself to find patterns or make predictions. Over time the human programmer can also tweak the model, including changing its parameters, to help push it toward more accurate results.
- Natural Language Processing (NLP): Natural language processing (NLP) refers to the branch of computer science—and more specifically, the branch of artificial intelligence or AI—concerned with giving computers the ability to understand text and spoken words in much the same way human beings can. NLP drives computer programs that translate text from one language to another, respond to spoken commands, and summarize large volumes of text rapidly—even in real time. There’s a good chance you’ve interacted with NLP in the form of voice-operated GPS systems, digital assistants, speech-to-text dictation software, customer service chatbots, and other consumer conveniences.
- Natural Language Understanding (NLU): Natural Language Understanding is a subset of research and development that comes from foundational elements of NLP. NLU analyzes what human language means, rather than simply what the individual words say, by looking at the connotations and implications in human communication like emotion, effort, intent, or goals behind a statement.
- Structured data: Highly specific data that is stored in a predefined format and is easier to manage and protect with legacy solutions. Prior to injecting chatbots with Large Language Models, chatbots were largely working off of structured data.
- Unstructured data: A compilation of many varied types of data that are stored in their native formats, such as images, audio, video, word processing files, emails, spreadsheets, and more. Large Language Models are able to pull from both structured and unstructured data to build their knowledge base.
- Fine Tuning: Adjusting and adapting a pre-trained model to perform specific tasks or to cater to a particular domain more effectively. For example, you can finetune a particular model for specific customer support use cases including tracking order status or providing product recommendations.
Common Large Language Models (LLMs)
- ChatGPT: ChatGPT is a large language model developed by OpenAI, based on the GPT (Generative Pre-trained Transformer) architecture. It is designed to engage in interactive and dynamic conversations with users, simulating human-like responses and generating text based on the input it receives. ChatGPT has been trained on a vast corpus of diverse text from the internet, allowing it to acquire a wide range of knowledge and linguistic patterns. It serves as a conversational AI system, enabling users to interact with it through chat interfaces and benefit from its natural language processing capabilities.
- Open AI: OpenAI is an artificial intelligence research organization and company that focuses on developing and promoting safe, ethical, and beneficial AI technologies. Founded in 2015, OpenAI’s mission is to ensure that artificial general intelligence (AGI) benefits all of humanity. It conducts cutting-edge research in various areas of AI and develops advanced models and systems, such as GPT-3 and ChatGPT. OpenAI also strives to make significant AI advancements accessible to the public and collaborates with the global community to address the societal impact and implications of AI.
- BERT: Stands for Bidirectional Encoder Representations from Transformers. It is an open source machine learning framework for natural language processing (NLP). BERT is designed to help computers understand the meaning of ambiguous language in text by using surrounding text to establish context. The BERT framework was pre-trained using text from Wikipedia and can be fine-tuned with question and answer datasets.
Customer Support and Chatbot-Specifc Terms
- Intent-Based: A type of chatbot that works well in simple, transactional inquiries (e.g. order status, shipping updates). This type of chatbot is typically built and trained on specific customer intents and their matching responses.
- Multi-Turn: The ability for a chatbot to answer multiple questions and decipher multiple intents within any given message or series of messages.
- Curated Knowledge Base: An extensive knowledge base consisting of datasets such as knowledge base content, product collateral, top-rated human resolutions, brand policies, and their own experience in the customer support and service space. This knowledge base is more targeted and customized when compared to the data that ChatGPT has access to on its own.
- At Simplr, we work with our partners to build a curated body of knowledge that ultimately feeds our generative AI chatbot versus relying on the full scope of content available in other large language models (like ChatGPT)
- Manipulation: When consumers are in a position to “trick the bot” into giving unapproved responses such as giving refunds or store credit outside of policy or engaging in unsafe conversations such as politics or religion.
- Hallucinations: The ability for LLMs to pull knowledge from wrong, non-traceable, or unapproved sources and appear convincingly right.
- Cognitive Paths™ : Cognitive Paths can model and guide a bot according to best customer experience standards and top-performing agent experiences. It is Simplr’s generative AI technology that has comprehensive safeguards to enhance customer service interactions. It reduces the overall volume of information available to the LLM-powered chatbot and eliminates the chance of hallucination, i.e., engaging in off-brand topics. The result is an extensive, curated knowledge base and a vastly more targeted and customized set of data than what ChatGPT has access to on its own. Cognitive Paths then directs the LLM to only pull information from certain datasets based on the nature of each unique customer interaction. The result is the ability to identify which stage of the conversation you are at and lead the bot conversation in a controllable, interpretable, and customizable manner while leveraging the effectiveness of generative AI without the hallucinations.
- Edge case: A problem or situation that occurs only at an extreme (maximum or minimum) operating parameter. With generative AI chatbots, it’s critical to test every possible edge case to ensure the chatbot is acting as intended.
Whether you’re seeking to grasp the basics or become a subject matter expert, we hope this glossary provides a foundation for the ever-evolving landscape of generative AI. Looking to take a deeper dive? Check out our comprehensive guide to all things GenAI.