5 AI Acronyms: What Do They Mean?

10 July 2023 | Luminance

Over the past six months, it’s been impossible to ignore the media frenzy around AI. But with terms such as GPT, ML, and NLP now regularly used across mainstream media, let’s get back to basics – what exactly do all these acronyms mean?

1. AI – Artificial Intelligence

Artificial Intelligence describes a machine’s ability to mimic human intelligence. That sounds simple enough but, in practice, AI combines a number of complex systems and processing algorithms in order to analyse patterns and features within data.

From self-driving vehicles on our roads to medical imaging in hospitals and fraud-detection software used by banks, AI is having a profound impact on our economy and society. Indeed, a recent report by Goldman Sachs indicated that AI is expected to raise global GDP by as much as 7% over a ten-year period. When it comes to Luminance, our AI uses multiple state-of-the-art techniques within the field of machine learning to read, understand, negotiate and analyse legal contracts.

2. ML – Machine Learning

Machine Learning is defined as a discipline within the field of AI that allows machines to automatically learn from data and past experiences to identify patterns and make predictions with minimal human intervention.

Luminance is unique in combining both supervised and unsupervised machine learning to provide the utmost legal rigour. For example, when taking a first pass review of any contract under negotiation, Luminance’s Traffic Light Analysis uses this powerful blend of machine learning to visually highlight which clauses are acceptable and which need further review.

3. NLP – Natural Language Processing

Natural Language Processing is, in effect, the process by which computers analyse spoken and written language. These applications automatically extract, classify, and label elements of text or voice data and assign a possible meaning to those elements.

Rather than examining text word-by-word, these models have a close understanding of what each word actually means in the context of a full sentence. In this way, Luminance can read and understand the written word much like the human brain, distinguishing between different uses of the same word depending on the context. For instance, let’s say a reviewer is looking for examples of witness statements containing the word ‘dog’. By conceptually understanding the word ‘dog’ and any variations of the word, Luminance would also be able to surface all similar terms such as ‘mongrel,’ ‘mutt’ or ‘German Shepherd.’

4. LLM – Large Language Model

A Large Language Model is a specific application of natural language processing. Moving beyond basic textual analysis, LLMs use advanced AI algorithms and gigantic datasets to understand, summarise and generate content. LLMs underpin many conversational AI chatbots and are also used to complete text-based tasks, such as text generation, content summary and even translation.

Luminance’s own LLM has been specifically trained on over 150 million verified legal documents and has obtained deep domain knowledge of the legal industry. This makes Luminance’s LLM ‘Legal-Grade’ – the most advanced language model accessible to lawyers today. Our ‘Ask Lumi’ chatbot is the latest application of this specialist LLM and allows users to ask questions about any contract and receive instant answers. It can also help with redrafting contracts in just a few clicks.

5. GPT – Generative Pre-Trained Transformer

If you’re already familiar with the term GPT, it’s more than likely you’ll have heard it used in conjunction with ChatGPT, a chatbot developed by OpenAI that can generate human-like responses to user prompts and questions. ChatGPT is so-called because it utilises Generative Pre-Trained Transformers: a set of algorithms, or deep-learning models, to generate new text, images or audio based on learnings from large, inputted volumes of training data.

In contrast to more generalist GPT technologies, Luminance’s Legal Pre-Trained Transformer (LPT) technology learns solely from legally verified documents and is able not only to produce content, but also to analyse and understand content – even that created or edited by third parties. When it comes to legal-specific knowledge, this effectively makes ChatGPT the equivalent of a well-read and entertaining dinner guest when compared to Luminance’s specialist LPT, which behaves more like someone who has studied law.

Interested in learning more about Luminance’s next-generation LPT? Click here to request a demo today.