Generative AI+ML tools
01-27-2024, 06:38 PM
Using AI models and machine learning to generate content is all the rage nowadays, so here are some resources if you want to get in on that.
Text: Enter a text prompt, get a text response back.
ChatGPT ( https://chat.openai.com/ ) The most basic, but also the most controversial. Ask it for advice or info like any search engine, converse with it like a person, or have it write some "original" content for you(donut steel). I've heard it helped people with programming problems but I haven't used it for that purpose yet.
Images: Enter text prompt, get an image back with the characteristics you specified.
Bing image creator ( https://www.bing.com/images/create) Very good general use image generator powered by the DALL-E 3 model, I've been getting the best results with this one. Each prompt will get you up to 4 images. Note that there are content filters that block NSFW content and anything involving the names of real life celebrities. Fictional characters appear to be fine though.
Stable Diffusion ( https://stablediffusionweb.com/ ) This web version isn't so good, but with a bit of effort you can set up an instance on your own machine and use a dataset that's closer to what you want.
Artflow.ai (https://app.artflow.ai) focused on creating faces
Audio: Reads a text prompt aloud (also known as Text To Speech or TTS)
Eleven Labs ( https://elevenlabs.io/ ) We've come a long way from Microsoft Sam. Most of the voices sound very natural
Text: Enter a text prompt, get a text response back.
ChatGPT ( https://chat.openai.com/ ) The most basic, but also the most controversial. Ask it for advice or info like any search engine, converse with it like a person, or have it write some "original" content for you(donut steel). I've heard it helped people with programming problems but I haven't used it for that purpose yet.
Images: Enter text prompt, get an image back with the characteristics you specified.
Bing image creator ( https://www.bing.com/images/create) Very good general use image generator powered by the DALL-E 3 model, I've been getting the best results with this one. Each prompt will get you up to 4 images. Note that there are content filters that block NSFW content and anything involving the names of real life celebrities. Fictional characters appear to be fine though.
Stable Diffusion ( https://stablediffusionweb.com/ ) This web version isn't so good, but with a bit of effort you can set up an instance on your own machine and use a dataset that's closer to what you want.
Artflow.ai (https://app.artflow.ai) focused on creating faces
Audio: Reads a text prompt aloud (also known as Text To Speech or TTS)
Eleven Labs ( https://elevenlabs.io/ ) We've come a long way from Microsoft Sam. Most of the voices sound very natural
