By
Published: 12 Feb 2024
There are many types of AI content generators with a variety of uses for consumers and businesses.
ChatGPT, a powerful AI chatbot, inspired a flurry of attention with its November 2022 release. The technology behind it — the GPT-3 language model — has existed for some time. But ChatGPT made the technology publicly available to nontechnical users and drew attention to all the ways AI can be used to generate content. Now, more than a year after its release, many AI content generators have been created for different use cases.
This has prompted questions about how the technology will change the nature of work. Some schools are banning the technology for fears of plagiarism and cheating. Lawyers are debating whether it infringes on copyright and other laws pertaining to the authenticity of digital media. President Joe Biden also passed an executive order in October 2023 that addresses the technology’s opportunities and risks in the workforce, education, consumer privacy and a range of other areas. Generative AI has the potential to change the way content is created.
What is AI-generated content?
AI-generated content — or generative AI — refers to the algorithms that can automatically create new content in any digital medium. Algorithms are trained on a large amount of data. Outputs are then returned based on that data and a comparatively little bit of user input. But the key is that content is new and generated automatically.
The most common example of a generative AI algorithm is ChatGPT. ChatGPT performs natural language processing and is based on the language model GPT-3. GPT-3 is trained on a large amount of human text from the internet and teaches the language model how to respond when interacting with users.
Other generative AI programs work in a similar way. They are trained to develop a body of knowledge and use that knowledge to create novel outputs.
Many commercial generative AI offerings are currently based on OpenAI’s generative AI tools, such as ChatGPT and Codex.
What are the different types of AI-generated content?
AI-generated content is not just limited to mimicking human writers. AI-generated content also exists in other media, such as the following:
Image.
Speech.
Video.
Music.
Code.
Despite the many types of content generative AI can create, the algorithms used to create it are often large language models such as GPT-3 and Bidirectional Encoder Representations from Transformers — also known as BERT.
How generative AI is used
These different media can be used in tandem to generate various content. They can be used in many fields, including the following:
Academia, to write papers and other long-form content.
Law, to write legal briefs.
Science, to automate and expedite drug discovery.
Art, to generate new works and unique content ideas.
Manufacturing, to expedite product development.
Digital marketing, to write copy and product descriptions, as well as draft social media posts.
Software development, to generate, remediate and summarize code.
Cybersecurity, to perform accelerated threat detection and develop malware.
AI content generators
There are AI content generator tools in every medium — some paid and some free. Many are based on similar technology and add features to address specific user needs. Below are some of the top content generators organized by content type.
Written
AI-Writer is an AI writing assistant that helps reword and remediate an existing piece of content, creates unique article drafts, provides citation lists, summarizes SEO competitors and creates SEO-optimized content from Google.
Gemini (formerly Bard) is Google’s conversational AI tool, powered by a language model of the same name. It is available in more than 230 countries and territories and functional in over 40 languages. Gemini competes with OpenAI’s ChatGPT and aims to improve the way people search for and retrieve information. One example of a use for Gemini is asking it for meal ideas based on what food is in the house. Another example is asking it to explain complex scientific concepts in a way a child could understand. Gemini, like ChatGPT, is multimodal, meaning it can handle images as well as text. Gemini integrates with other Google services such as Gmail and Google Docs.
ChatGPT is a machine learning-powered chatbot trained to engage in realistic dialogue. The chatbot can debug code, challenge incorrect premises and refuse to respond to certain requests. The model produces convincing content but does struggle with factual accuracy. It was trained using reinforcement learning and is informed by previous OpenAI projects such as GPT-3 and Codex. It also integrates with image models to generate images from a text prompt.
ChatGPT can help users troubleshoot and generate computer code.
ChatSonic is a dialogue-focused content generator from Writesonic built on top of GPT-4 with added features. Users can choose from 16 different personas to chat with, including poets and accountants. Users can also interact with ChatSonic via voice and choose to have ChatSonic respond audibly. ChatSonic is connected to Google’s knowledge graph and includes up-to-date information from the internet.
Claude is a conversational AI tool by Anthropic, an AI safety and research lab. Claude is in beta and is integrated into Slack to summarize threads and answer questions. It provides a principal-based approach to performing chatbot functions and can handle adversarial questions. Anthropic was founded by two former OpenAI employees.
Docs is Google’s cloud-based, collaborative word processor with AI features to generate, summarize and brainstorm text. Docs is integrated with Google Bard as well.
Jasper AI is a long-form AI copywriting tool and article generator that includes more than 50 content generation templates in 25 global languages. Jasper’s user interface is simple to use and lets content creators specify SEO keywords and tone of voice in the prompt.
Magic Write is an AI text generator for Canva Docs. Users can prompt magic write to brainstorm, generate outlines and generate content ideas. It was developed using GPT-3.
Perplexity AI is an AI-powered search engine. It includes many of the same natural language functions that ChatGPT has to offer. A unique feature of Perplexity is that it can provide sources to back up the answers that it has generated. It functions more as a search engine than an original writer, but still generates original content.
Rytr is an AI content generator that lets content creators specify a content use case, tone of voice and keywords. It also comes with plugins for WordPress.
Spellbook by Rally is writing software designed to help lawyers with legal drafting. Spellbook can draft new contract clauses, list common negotiation points based on the contract and create contract summaries.
Wordtune is an AI tool for AI21 Labs that can perform a variety of general language tasks, such as rewriting sentences and citing sources.
Image
Craiyon, originally called DALL-E mini, was developed by researchers at Google and Hugging Face. It produces a batch of AI-generated images in response to a text prompt. The image quality is generally lower than that of Midjourney, Stable Diffusion and DALL-E. The images often appear blurry and unrealistic.
DALL-E is OpenAI’s image generator that creates images and art from a simple text prompt. DALL-E can also make realistic edits to images using a caption. It can remove components and change image texture, for example. It can also create several variations of an existing image based on the original. DALL-E uses diffusion, which starts with a random pattern of dots and creates an image from it.
Midjourney is an image generator hosted on a Discord server. Users can manipulate the generator via Discord bot commands or through a web app. It generates realistic-looking images based on text prompts. Some users have noted that Midjourney output resembles paintings rather than photographs. Users are encouraged to post their creations on designated X — formerly known as Twitter — and Reddit pages to help train the system.
Stable Diffusion is Stability AI’s generator that can create photorealistic images from an inputted text. It can also be used to fill in missing or deteriorated pieces of art and has an easy-to-use in-browser function through DreamStudio. Stable Diffusion uses diffusion to create images.
Music
Amper music is a generator that creates music from prerecorded samples. The software can be used to match music to video, where users can specify the duration of music and where in the music a climax should fall.
Dadabots is a generative neural network that creates a constant live stream of artificial music in different genres and raw audio neural networks that imitate bands. For example, its infinite bass solo video is a 24/7 stream of emulated bass noises.
Harmonai is a community-driven organization that creates open source tools for generating novel sounds based on diffusion, a process that takes white noise and manipulates it into audio samples automatically. Harmonai aims to develop open source audio synthesis and manipulation tools. Harmonai released Dance Diffusion, which has six publicly available music generation models — each trained on different sets of data sets.
MusicGen is a Hugging Face Space that lets users generate audio from a description, an example file or a sound clip recorded directly from the device’s microphone.
Soundraw is an AI music generator that can be used to generate royalty-free background music. Users can choose the tempo, mood, genre and theme of the song.
Code
CodeWP is an AI code generator specifically for WordPress. It works with JavaScript, PHP, WooCommerce, Breakdance, Oxygen and Regex conditions. CodeWP helps WordPress users by minimizing the time spent on development. Like other content generators, CodeWP works based on small text prompts.
CodeWhisperer is an AI coding companion from Amazon that provides AWS users with code recommendations based on existing code and comments. The tool can also highlight code that might create a security vulnerability and suggest changes.
GitHub Copilot is an AI-powered code generator that generates code from natural language prompts in the code comments. For example, a coder could write “design a website landing page” and it would produce the appropriate code. It is free to use for verified students, teachers and open source project maintainers. Others can use Copilot with a paid subscription or through a one-time 30-day free trial. CoPilot also has an Enterprise tier that corporate developers use as a pair programmer with internal source code.
Tabnine is a coding assistant that predicts and generates a programmer’s next lines of code based on syntax. Tabnine makes suggestions based on a coder’s style and programming patterns. It works with a variety of languages including Ruby, Python, Rust and Go.
Other
Absci is a drug creation company that designs antibodies from scratch using AI. Their Integrated Drug Creation platform designs antibodies without prior training on an existing target. This speeds up the process of drug discovery, enables the design of new drugs and enables personalized medicine. The scientist prompts the model with a target protein and the program generates an antibody in response.
Bing is Microsoft’s AI-enabled search engine with chatbot functions. Users can query it the way they would a normal search engine, and it will return fully synthesized answers instead of the standard list of links.
Creo is a computer-aided design system that uses generative design, enabling expedited design of physical objects. It also optimizes designs based on material and manufacturing requirements.
DuckAssist is an AI feature in the DuckDuckGo search engine that generates answers using auto-summaries of Wikipedia pages. It aims to limit natural language models’ tendency to hallucinate answers and provide incorrect information.
Duet AI is a set of AI features for Google’s enterprise collaboration suite, Workspace Enterprise. Duet embeds AI features in several of Workspace’s tools. Google Slides can produce AI images, audio and video for presentations. Google Sheets can automatically derive insights from data and generate formulas using AI. Gmail can draft and generate email replies automatically.
Grok is xAI’s — Elon Musk’s AI startup — version of ChatGPT. It has internet browsing capability and can answer users based on up-to-date information on the web. The tool is designed to answer questions that other AI systems might typically avoid and have a more humorous tone.
Pi is a chatbot designed to serve as the user’s personal assistant. Inflection AI launched the tool. It is designed to offer personal advice and support. It comes from LinkedIn cofounder Reid Hoffman and DeepMind cofounder Mustafa Suleyman’s company Inflection AI.
Supermeme.ai is an AI-powered meme generator. Users can input several hundred characters of text and the program generates a meme based on it. AI automatically chooses a template and writes a caption. Users can also generate memes based on a theme. The memes follow a classic “bottom text” format.
You.com is an AI-powered search engine with chatbot functions. Users query You.com, and the search engine returns a summary of search results in a conversational tone. You.com also provides citations for its responses. But it is not limited to text — it also features an image generator and code generator.
While automatically generating content has its benefits, it’s also fraught with risk and uncertainty. Read up on some of the pros and cons of AI-generated content.
Ben Lutkevich is a technical features writer for WhatIs.com, where he writes technology explainers and definitions.
Next Steps
AI washing explained: Everything you need to know
Top AI podcasts to listen to
Differences between conversational AI and generative AI
An explanation of the different types of AI
Will AI replace technical writers? One user’s experience
Related Resources
Dig Deeper on Data analytics and AI
