Four AI concepts marketers need to know

- Before we dive into the AI tools and how you might use them for your brand, it's a good idea to understand some key AI concepts. Let's look at 4, natural language processing, or NLP, computer vision, multimodal learning, and generative AI. I'm going to start with generative AI, since it's a term you've been hearing about a lot. Generative AI describes what happens when a machine is able to produce something original based on a written request. A simple way to think of it is prompt to create. You describe what you want, and the AI brings it to life. How does it do that? Well, generative AI uses natural language processing, or NLP to- in quotations- "understand" the connection between words, so it can predict what might come next. You see, in order to provide a human-sounding output, NLP needs to be trained on vast amounts of data, and that's where large language models, or LLMs, come in. LLMs read or process data from websites, digital books, social media posts, blogs, media outlets, Wikipedia, and many other public sources of online text. Once an NLP has been trained, it can use its knowledge to offer a good guess of what may come next in a sentence or paragraph and craft a verbal response that makes sense to you and me. In the same way that NLP lets a machine understand words, computer vision is all about the way an AI sees the world, and it does that in a very different way from people. When you see a chair, it doesn't matter if it's upholstered, made of wood, or even upside down, you know it's a chair. Machines need to be trained on many different images of chairs of all shapes and sizes in order to distinguish them, because they don't possess the reasoning abilities we have, and that can lead to AI mistakes. Until recently, NLP and computer vision operated as separate domains, but breakthroughs in multimodal learning let machines combine several data types to produce creative results. Now, that paved the way for text-to-image design, and will soon alter how you create videos, music, synthetic voices, 3D scenes, and a lot more. Ultimately, it all comes down to the quality of your prompt and expressing what you want in language a machine can grasp, which is why prompt engineering is a good skill for marketers to have.