Whether we like it or not, AI has taken the world by storm; arguably disrupting our education system, work force and our social lives. The conversation around AI and its functions had introduced a lexicon with phrases and buzzwords that when used, continue to mold how AI is being perceived by the wider community.
As students, and more importantly, student journalists, becoming vigilant and critical of AI and the language surrounding it is incredibly important. Terms like Large Language Models or Bias have been floating around the casual and academic space, and if anyone else was just as confused at what these terms mean, you’re not alone. At the Gauntlet, informing our student community is the heart of our operations at the most accessible level possible; so to manage the AI jargon, this is a small guide to the most common terms and what it means in the context of AI.
Bias in AI
In the context of AI, bias appears when a result from an AI platform is exhibiting prejudice or discrimination based on the data it was trained on.
Deep learning
Deep Learning is similar to Machine Learning where it picks up patterns from various sets of data. By using networks to analyze the data, it creates a hierarchy of patterns that is based on a prompt
Foundation models
A Foundation Model is a large information model and AI platforms are trained on these models to complete a task or respond to a prompt.
Generative AI
Generative AI’s most basic definition is that it has the ability to generate new content from existing data and patterns that were used as reference.
Gigawatts
According to Merriam Webster, a gigawatt is “a unit of power equal to one billion watts.”
Hallucination
Many students who’ve interacted with AI or chatbots have learned to recognize a hallucination. Hallucinations are considered a misleading aspect of AI where a result will include fabricated information.
Large Language Models
Large Language Models is a new term that has entered the conversation around AI and its definition is still debated or misunderstood by many. The system is trained on textual data that can allow Generative AI platforms to create their own content.
Machine Learning
Like Large Language Models, Machine Learning allows for AI systems to recognize patterns to produce results that replicate or predict patterns that are successful. Machine learning is the base for many predictive AI models.
Token
A token is a broken down version of a longer text or word into a numerical form, and Large Language Models use tokens to identify words and generate sentences.
AI’s purpose in the academic sphere
Why do these terms matter? Simply to understand how AI works. Instead of starting conversations off with baseless assumptions about AI, everyone should do their due diligence in informing themselves of how AI works and what it means for an individual’s ability to make a critical observation, produce art, process emotions and more. In a period of mistrust and disinformation AI seems to answer questions in a simple and definitive way. There’s no hemming and hawing from an indecisive colleague or even a wait time to get an answer to a simple question. And maybe, that’s what makes AI appealing and reliable on various levels.
The current use of AI aims to fill the gap that was created by the overwhelming uncertainty that we as a generation are feeling. Amidst political unrest and workforce anxiety; ChatGPT becomes substitute therapists, Google Gemini can answer any question thrown into the search bar and Dall-E has the ability to produce a picture without even picking up a camera.
Because AI made a seamless transition into everyone’s lives we are forced to reckon with the consequences that are associated with it. Which begs the question why even use it at all? Perhaps instead of relying on AI to give a simple answer; people should do the work and read the texts that trained these systems or attempt to draw a really bad sketch. Although the answer to reject AI is easier said than done, technology that is used at the post-secondary stage is developing rapidly and AI, for better or worse, is integral to that growth. University of Calgary students are now entering a new semester where AI is now part of the course outline despite the university itself not having a strict AI policy. During this period of technological exploration, keep the definitions above in mind and remember the works and stories that AI is trained to reproduce. The most that students can do is remain attentive when navigating the technological conundrum that is AI.
