Decoding AI Jargon (for normal people)

If you haven’t already, it’s time to come to terms with the fact that AI will be taking centre stage for the foreseeable future.

Particularly for us marketers.

A recent study by Hubspot found that 81% of people in our industry say AI is effective at assisting them in their role.

On top of that, 56% of marketers who use generative AI for content creation say their content performs better than content created without it.

If you’re a late adopter, that’s fine. It’s a crazy world we live in.

And with new technology coming at us faster than ever before, it’s okay to be overwhelmed (and confused, because wtf are Neural Networks and why are they smarter than me?)

But PUT DOWN THE GPT prompts. You’re going to need to brush up on your lingo before anything.

And when I say you, I mean us, because girl… I just got here, too.

Lucky for us, Asa Hiken and Garett Sloane of AdAge created a glossary that’s going to whip us into shape in no time.

According to them, those of us 'eager' (don’t know if that’s what I’d call it, but alright) to incorporate AI into our strategies 'need to first learn the key terms that articulate the space.'

And this vocabulary, they explain, is like 'a stack of nesting dolls' that endlessly bends back on itself, requiring knowledge of the prior concepts.

Great! Sounds simple enough. If simple is another word for nightmare.

'How else is one supposed to understand LLMs if they don’t already have neural networks down? Worse yet, how embarrassing it might be to discuss LaMDA when really you meant to talk about LLaMA…'

Yeah, sure buddy, whatever you say.

So, without further ado, let’s jump into the glossary of terms we need to know in our current dystopian reality.

(The original article is subject to change, as they will be updating it as new terms arise. If you’re here well after the publishing date, you may want to double back and check out what else there is to know.)

Let’s start with the main thing you should know:

Artificial Intelligence. This is the general term for the set of technologies that use advanced computing to perform tasks. There are many developments that belong under the umbrella of AI, like generative AI. While it is a subset, AI doesn’t always refer to tools that generate content. The history of AI dates back to the 20th century. But advancements beginning in the 1960s led to the capabilities we have today.

Chatbots. These are programs that run inside messaging apps/ websites that help consumers perform simple tasks, like the ones you see on Facebook Messenger. The recent boom in generative AI bots like ChatGPT mean these services are becoming more human-sounding. Greeeeeat.

Speaking of which, let's talk about ChatGPT. I’m sure we all know what this is by now. But just in case, ChatGPT is a generative AI bot created by OpenAI. It uses GPT technology (see below) to produce an output based on an input (which is what you ask it). The platform has been trained using vast quantities of text from websites, articles, books etc.

Clusters. This is a fancy word for a group of people—or anything—sharing a common characteristic. AI programs can identify clusters within mountains of data, uncovering patterns that us mere mortals alone can’t perceive or connections that we wouldn't draw. (Ok settle down we’re not that stupid.) Clusters can lead to developing audiences or segments for marketing purposes. They help us create a group of people with common traits and target them with ads.

DALL-E. No, not Mister Worldwide. This is another generative AI bot created by OpenAI, but this one generates images. Like ChatGPT, a user queries the bot with an input (e.g. 'make an image of Ronald McDonald swimming in a river of fire') and out comes a corresponding image. The latest versions can produce an array of images in a range of styles (e.g. 'Ronald McDonald swimming in a river of fire, in the style of Norman Rockwell').

Deep Learning. A more advanced branch of machine learning, this is where a computer teaches itself with only minimal amounts of programming. With deep learning, marketers can make the most use of data and apply it to make predictions about consumer behaviour. Not scary at all.

Generative AI. As the name suggests, this is a form of AI that generates content. Depending on the specific platform’s capabilities, this content can be images, videos, text, audio, as well as other mediums. The platforms that have recently become popular generate their content from a user’s input, such as 'Produce an image of X' or 'What is the answer to Y?'

GPT. This stands for Generative Pre-trained Transformer. No, not like the Decepticons (I wish). Instead, it's a generative AI model that is trained on reams of data to produce an output.

Image recognition. AI looks for patterns in images. Machines can analyse many more images than humans. And with machine learning, they can identify what's in the images and reveal patterns that people would never detect (we get it, humans ain't sh*t.) As a brand, you could use image recognition technology to find every photo online in which your logos appear. That could help you locate your most loyal customers and tease out other actionable marketing insights.

LLM. This is the Large Language Model, a class of AI technology whose primary function is to create human-sounding text. The platforms are trained on reams of data and built on intricate neural networks that enable them to make predictions and generate corresponding text. OpenAI’s GPT model is an LLM.

Machine learning. Basically, this is when a machine teaches itself with minimal programming needed. Machine learning can be helpful in direct marketing and email marketing. By ingesting large amounts of consumer data, you can use it to determine things like the best times to send emails, etc.

NLP. This stands for Natural Language Processing. It's the technology that enables machines to interpret what people are saying in words or in text. Sophisticated AI can decipher speech, not just understanding the words but the context. It can even detect sarcasm and other subtle human tones.

Neural Networks. Or, the thing I said was smarter than me. These are Artificial Intelligence programs modelled after the human brain. They incorporate deep learning and natural language processing to perform functions like recognising handwriting and faces in photos.

Weak/ Narrow AI. I find this so funny because we’re already, like, calling lesser forms of AI names. I can imagine in the future, this will be a robot slur. It means AI that's limited to specific tasks. Basically, all the AI found in marketing is weak. Apparently, MOST AI, from virtual assistants to self-driving cars, is considered weak. The next evolution is artificial general intelligence. And this will bring us closer to the long-anticipated, futuristic robots that could outperform any human at any task.

Right, yeah because that’s totally what we all want (help ????)

Anyway, if you made it to the end of this, I’m proud of you! You deserve a treat.

And remember, if you knew all of this already, it doesn’t make you better than any of us, nerd. We are all in this together. No, seriously, we need to stick together for when the robots take over.