Last Updated on January 11, 2024 by Alex Birkett
Robots are writing copy, making images, having conversations.
From the popular large language models created by OpenAI to the more recent ChatGPT (also by OpenAI) and open source models like Stable Diffusion, this stuff is getting wild.
I, for one, have embraced AI content writing tools. They help me break through writer’s block, repurpose content, and speed up the content creation process.
Generative AI models are also being built into basically every SaaS product you use on a daily bases.
However, there’s a lot of confusion and cacophony out there about what these generative models actually do.
This post is my attempt at a comprehensive guide and explanation for what is being called “generative AI” and all of the tools associated with it.
This space is moving incredibly fast, though, so I’ll update the article as things change.
- What is Generative AI?
- How Does Generative AI Work?
- Why is Generative AI Important?
- The Response to Generative AI is Mixed
- The Best Generative AI Tools
- Conclusion
Note: I’m including a few affiliate links in this piece, primarily to Jasper. I love the product, use it regularly, and get a small kickback if you use my link and purchase. It doesn’t affect the tone or opinions in my content.
What is Generative AI?
The definition of generative AI is evolving, but the consensus is that it’s a class of deep learning model that is capable of producing new outputs (images, text, video, etc.).
- Generative AI is a type of artificial intelligence that focuses on creating new content, rather than recognizing patterns or understanding language.
- Generative AI has enabled machines to generate output that is far more creative and complex than ever before, and it’s being used in many fields, from copywriting to marketing.
- Generative AI is a type of artificial intelligence that uses algorithms and machine learning to generate new outputs based on a set of inputs.
This can either be done by training an algorithm on existing data or by creating a generative model from scratch.
The final output generated by generative AI can range from simple text-based content to images, audio files, and videos.
Examples of generative AI include using natural language processing (NLP) systems to generate new texts based on an existing dataset or using deep learning models to create realistic images from scratch. It can also be used for AI research (such is the speciality of Perplexity AI).
Take, for example, this article. I outlined most of the piece using Jasper, which saved me probably 1-2 hours in the initial writing phase.
For me, the owner of a content marketing agency, this saves me a ton of time while allowing me to maintain creative control of the content.
Or, take this image I generated using Midjourney. I’m throwing a Christmas party, so I just added a little prompt (“people in suits and santa hats drinking cocktails, etc. etc.”) and it generated this beautiful work of art:
There are plenty of examples of AI generated videos, too, though I’ve explored that domain less than art and text.
How Does Generative AI Work?
Generative AI works by taking existing data as input and then generating new outputs based on this data.
It starts by analyzing the input data in order to identify patterns in the data set.
This allows the system to understand the structure of the input data and how it relates to other pieces of information in the dataset. Once the patterns have been identified, the system can then generate new outputs based on these patterns.
For example, if you feed a generative AI system a large dataset composed of customer reviews for a product, it could be trained to generate new reviews for similar products.
A lot of the text generation algorithms are likened to fancy auto-complete functionality. Given a ton of past data (from these tools scraping a lot of content on the web and creating text models), it can predict the next word, sentence, and paragraph (in combination with the prompts and instructions you give it).
There are many different models, however, so I want to make sure to note that not all of them work the same way.
Several AI copywriting tools use different models.
Some are built directly on top of GPT-3, but others have proprietary NLP machine learning models of their own. Some, like Jasper, test many different models for accuracy and utility.
Why is Generative AI Important?
Generative AI has become increasingly important for many industries due to its ability to quickly generate high-quality content at scale.
For instance, copywriters can use generative AI systems to quickly write hundreds of blog posts with minimal human intervention; marketers can use them to automatically create ads tailored towards specific audiences; and content writers can use them to produce engaging articles with minimal effort.
Speed is a massive advantage, especially for startups. And there’s never been a force multiplier like AI when it comes to speed, at least in my lifetime.
Furthermore, generative AI systems are being used in many other areas such as medical research and drug discovery where they are able to uncover insights that would otherwise go unnoticed by humans alone.
With the pace of AI accelerating rapidly, however, there are some very exciting developments as well as some worrying concerns.
Fundamental questions like “what is art,” “what is creativity,” and “what does it mean to be human” are coming up.
In the meantime, we can use it to analyze complex data points (particularly qualitative) and speed up time to insights. We can generate content faster. We can streamline multiple tasks and automate business processes.
For now, neural networks and AI algorithms are simply important because they are useful (and will only get more useful).
The Response to Generative AI is Mixed
The benefits to many seem obvious, but the response has been mixed.
Many people, rightly, are skeptics. How can a computer write poetry, novels, or even blog posts with the soul of a human?
The short answer is, at least for now, they can’t. These tools are best looked at as assistants to the creative process, not replacements.
One of the most thoughtful takes I’ve seen is by Jay Acunzo, someone I really respect in the field of content marketing:
In my opinion, a lot of skepticism has stemmed from fear. And my friend Tommy Walker, a fellow creative and writer, expressed this well:
It does make one nervous and contemplative, though. What happens when AI generates better songs, poetry, and videos than the average person could?
Perhaps the scarcity, then, of human creativity will increase its value?
There are a lot of unknowns at this point. All I know is that AI text generation and image generation has been a huge boost to my creativity, not a hindrance or a replacement.
And I’ll probably always create art as a mode of self-expression. As for the future value and utility of that creativity, though, that’s up in the air.
The 4 Best Generative AI Tools
There are many generative AI tools, and more come out every day. Some are core models, and some are UI layers built on top of the core AI models.
Here are some of my favorites:
- Jasper
- OpenAI
- Stable Diffusion
- Midjourney
Note, too, that many products will begin incorporating generative AI into their core products. The distinction between an AI-company and a non-AI-company will be less clear.
For example, cold email outreach software companies will very likely incorporate the ability to generate personalized email copy.
1. Jasper
Jasper was one of the first AI copywriting tools I ever used, and to this day, it’s my favorite one.
The team is just incredible, the UI is super easy to use, and the product itself gets better and better every day.
When they launched, it was alright. You could use some basic templates and generate decent copy.
Now, through their Boss Mode editor and powerful commands, you can write pretty much anything. They also have a Chrome extension and just acquired Outwrite, so you can generate AI text anywhere.
They also have an AI image generator that is pretty solid.
2. OpenAI
OpenAI is an open-source Artificial Intelligence research lab created by Elon Musk, Sam Altman, Ilya Sutskever, Greg Brockman, Wojciech Zaremba, and John Schulman.
Their goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by traditional corporate concerns.
OpenAI’s main product is their GPT-3 AI natural language processing system, which can generate incredibly realistic-sounding text on virtually any topic.
They also launched DALL-E, the image generation model, as well as recently ChatGPT.
OpenAI is at the forefront of generative AI and they’ve been the leading technological pioneers in the space.
3. Stable Diffusion
Stable Diffusion is an open source AI model specializing in image and art generation. They use a variant of Generative Adversarial Networks (GANs) to create realistic images and digital art.
The Stable Diffusion model was trained on millions of images from the ImageNet dataset, allowing it to generate any kind of image you can think of. It also supports custom datasets so you can train your own AI model if you want to.
4. Midjourney
Midjourney is another AI image generation tool – my personal favorite one.
Like the others, you enter a text prompt and it returns an image. Actually, Midjourney sends you back four output variations and you can choose to iterate on any of them or upscale them in quality.
You can also upload parent images (i.e. images or photos you already have) to act as guidelines and inputs on image creation.
I’ve seen this tool be used for super creative use cases, like my friend who writes fantasy novels using it to visualize the world she’s writing about.
Further reading on Generative AI Tools
I’ve written a ton of content comparing and reviewing specific AI tools in this space. If you’re interested, check out the following:
- Categorical reviews:
- Product specific reviews:
Conclusion
Generative AI is an incredibly powerful tool that can be used in countless different applications ranging from copywriting and marketing all the way through medical research and drug discovery.
It has enabled machines to generate highly creative output at unprecedented levels of accuracy and speed – something which was impossible just a few short years ago!
As more organizations begin leveraging this technology for their own needs, it’s clear that generative AI will continue playing an instrumental role in shaping our world for years to come!