This post was contributed by Provar’s Chief Strategy Officer, Richard Clark. For more blog posts contributed by Provar’s executive leadership team, be sure to check back!
You can barely use social media, watch TV and streams, or read a newspaper now without seeing a mention of artificial intelligence (AI). Whether espousing AI as a tool of the devil or a super being, it’s important to understand some of the new terms that are being used. By no means is this complete! This list is meant to be fun and digestible, not as an all-encompassing reference guide.
I’ve summarized below my A-Z list of some of the most popular terms that have entered common usage simply to help everyone understand to a common level, what they mean, and why they’re important for Generative AI solutions like ChatGPT and the wider AI revolution currently underway.
AI is the broad concept of having machines think and act like humans. Generative AI is a specific type of AI.
Any large language model (LLM) is limited to a set of data on which it bases its answers. Like humans it can only repeat what it has learned, but like humans it also introduces subtle variances (see Hallucination). If we feed an LLM based on the National Enquirer, we’d get different results than if we used The New York Times (well, you’d hope so anyway).
A conversational AI developed by OpenAI to allow users to interact with its LLMs through a chatbot interface. The GPT element stands for Generative Pre-trained Transformer, which is a fancy way of saying that it’s an LLM using historical data sets (pre-trained) and uses a unique algorithm for generating human-like responses (using natural language processing). As OpenAI increases the data set on which their LLM is trained, they release version updates such as 3, 3.5, and 4 which reflect the currency of the information used. Despite the lag in information currency, these models can be augmented through Grounding to increase the accuracy and relevance of results.
Broadly used within the domain of Machine Learning, which includes the method in AI for teaching computers to process data much like the human brain. It’s the fundamental method that enables AIs to more accurately recognize pictures, text, and sounds, including when only incomplete information is available. Using pre-classified examples, the model learns to recognize patterns and apply the knowledge when it comes up against new content it has never encountered before. The word “deep” applies to the multiple layers of the neural network used.
It’s vital everyone understands AIs are fallible, just like us. To measure this, Error Metrics are a mechanism for benchmarking how well an AI is performing. When you increase the Variance in an AI model, you also increase the errors.
The process of adjusting the weighting of parameters (or more specifically Hyperparameters) when using an LLM to optimize the results. If I want an AI to suggest something unique, I would increase its Variance. If I want a precise, textbook, consistent response then I would lower it. You can also fine-tune responses by feeding back Generative AI responses and asking it to improve its answer to take into account additional Grounding instructions.
I wanted to cover GPT but Grounding is more important to understand. Grounding is a mechanism for reducing Hallucinations in the AI responses through the use of additional external data sets.
It’s one of many mechanisms to fine-tune and train the model on your specific requirements. Salesforce’s new AI Cloud is very interesting with a built-in capability to augment publicly available LLMs with customer data sources.
For Salesforce customers it would include using the data in your systems to give more narrow and specific results. What works for my competitor on their product sales as a next best action may be entirely different to me based on my product’s different capabilities, customer demographic, and historical success and failures.
You can also Ground a model through human feedback on the quality of responses.
Occurs in AI when the output from a generative AI in particular comes to a false result and creates content that is inconsistent with reality. In my experience, if I ask GPT to create a unit test following best practices it creates unnecessary teardown functions and annotations that don’t exist, but in reflection could be a good idea in the future!
While very interesting for creativity purposes and innovation, we generally think of Hallucinations as undesirable outcomes, and they can indicate a problem in the model or insufficient prompting and Grounding to accomplish the task requested.
Prompt chaining is a useful mechanism to get the model to re-evaluate its own previous answers, find errors, and fix them. It’s very important to understand Hallucination and not blindly trust the AI output without validation.
The process where we expose an AI model trained on one set of data to a new scenario or data set it hasn’t seen before and allow it to apply what it’s learned on its trained data to make predictions or classifications based on its previous knowledge.
The probability of two or more events occurring simultaneously. In AI it can be used to determine the dependency between events or variables.
K-means, K-nearest, K-fold
This is the point where you go read a proper machine learning data science paper. Just be aware these terms exist. Next.
Large Language Model (LLM)
Of course we all know what an LLM is, right? In essence, an LLM is a type of AI trained on a lot of text data. Using this training data, it can be engaged in conversation to generate human-sounding responses. Different LLMs are specialized at different things, like writing poetry, generating code, or answering questions, while others like ChatGPT attempt to provide answers for a broad set of needs.
Before generative AI hit the headlines, most AI applications in software were specifically relying upon Machine Learning (but most still can’t tell you how they use it, hmm…). Despite deliberate miscommunication by vendors, Machine Learning specifically focuses on the classification of existing data only. This is important for the pre-trained aspects of GPT tools but not the same as Machine Learning does not make predictions or have the capacity for Hallucination or the creation of new content. Machine Learning can provide probabilities. If you show it a picture of a dog and have trained it only on cat pictures, it will probably say (depending on the dog) it’s 20% likely to be a cat when a human can say categorically (most of the time). Remember this when using your camera phone to identify edible plants when foraging. Yes, it could still be deadly nightshade you’re about to put into that salad! Do you feel lucky?
Natural Language Processing (NLP)
NLP refers to the branch of AI dedicated to understanding human language and its real use, not just perfect grammatical inputs. It includes both text and voice inputs and is used by Generative AI both in the processing of requests and the generation of results. You can ask a Generative AI to make its responses more or less formal, or you can ask it to write for a specific audience or age category.
Can occur when an AI model has a too-narrow set of training data and as a result, becomes too specialized and less reliable at handling Inference. You could have a Machine Learning system only trained on cats, which means it’s much worse at understanding why a dog isn’t a cat than one trained on both cats and dogs. Don’t forget to use foxes either … I’m sure they’re really in between the two!
Prompting, or Prompt Engineering, is an AI technique used to provide more detailed instructions to the LLM to guide the output for your specific requirement. This can include informing the AI what its expertise is “As A Quality Engineer.” Related to this, templating inputs to AIs is a useful way to structure the repeatability of responses and reduce the verbosity of typed requests, or limit input for specific topics.
Chain of Thought Prompting allows you to tell the AI you’re going to provide input through multiple steps and wish to receive responses in multiple outputs too. You can also use Prompt Chaining to take the output of one request and feed it back into the AI with a request to use that output to perform further analysis or generation. Example: Create me an Apex trigger to XYZ -> Using this Apex class create a Unit Test using best practices -> Create me a Salesforce Flow that passes this Apex Unit Test.
A mechanism for making AI models smaller and faster through the approximation of numbers, which reduces the accuracy with the benefit of using less memory and processing resources. Expect to see more use of this term as Generative AI distributes its processing away from large server farms onto smartwatches, phones, and on-device use cases.
Reinforcement learning is a process for training AI models to improve their results through feedback about the output provided. This can be used for Grounding but more generally it’s the feedback that can be collected from human input to improve the results of more general Machine Learning. When I show my cat detector a picture of a dog and it says it’s 30% likely to be a cat, I need to tell it, no, this is a dog, for it to learn what dogs are. Eventually, it will tell me it’s 90% a dog and not 10% likely to be a cat.
A paradigm for teaching a model using examples with known answers. It’s used heavily in image recognition, translation, and predictions to improve accuracy before introducing unseen data. See Inference.
A type of deep learning model and incredibly useful for NLP. Transformers help tokenize the input received, apply the model’s understanding of the meaning and requests, and create the appropriate output unique to that combination of request, understanding and type of response required. Transformers are able to process based on a series of interactions (a conversation) rather than only on a single question. This is why when we both ask ChatGPT the same question we may get different responses depending on our prior conversations in that session.
Unsurprisingly, Unsupervised Learning is the process of letting AI find hidden patterns in your data without specific guidance on what’s right or wrong. This allows teams using AI to discover new patterns and correlations in data that may have not been known previously. One day we may truly know if a butterfly flapping its wings can cause a storm on the other side of the world.
In Machine Learning terms, validation is used to check how well the model is performing both during and after the training process. The model is tested using a new, previously unseen, data set, to validate it is learning and not just repeating answers to questions it already knows. It’s a useful mechanism to validate if Overfitting is occurring.
Weight Initialization is the process of setting starting hyperparameters prior to training the model. Effective weighting can impact the convergence, speed, and efficiency of LLMs. Each LLM is likely to have different Weight Initialization, which can affect their specific specialization along with the training data used.
eXplainable AI (XAI)
Provides insight into what has influenced the model’s results, which is important for understanding and trusting its results. It can help understand reasons for bias for example, or opportunities to improve results to be more accurate. Transparency of results is a major requirement for the ethical use of AI. When discussing with app vendors how they use AI in their products and where their data is coming from it is unacceptable, and unethical, for them to say, “I can’t tell you, it’s intellectual property”.
You Only Look Once (YOLO)
YOLO relates to real-time object detection. It’s an algorithm that uses a single pass through an image or video rapidly to allow objects to be identified in a single scan. It’s commonly used in self-driving cars, security cameras, and robotics. It has absolutely nothing to do with reckless behavior or social media memes.
Zone of Proximal Development (ZPD)
Ah, you thought I wouldn’t have a Z didn’t you? ZPD is an educational concept. Just like children going through school, ZPD is the fancy name for the process of training a model progressively more difficult tasks, so it can improve its ability to learn.
Interested in learning more about how Provar is using intelligent capabilities in its solutions? Connect with us today!