Generative AI is about to reprice the entertainment and art industry. The resources to create the most compelling imagery, video, and writing will only be constrained by one's imagination, not their wallets.
@aibreakfast
-
Testing Stable Diffusion 2’s image generation capabilities
By
–
Testing out Stable Diffusion 2 on @playground_ai – quite impressed with the level of detail from simple prompts.
-
User Experience with Overdub AI Voice Technology
By
–
Haven't used Overdub, but just checked it out and will try it. I think it's a technology that will get us to 95% accuracy really quick, but take some minor sound editing (for pauses and inflection) to make it truly passable as a natural sounding voice. But it's damn close.
-
Clone Your Voice Using Resemble AI Text-to-Speech
By
–
You can clone your own voice into a text-to-speech AI using Resemble AI, it's free to try and it's pretty wild. (this is not an affiliate link or anything, just sick af)
-
Debate: Best AI Image Generator Model?
By
–
Settle an internal debate: Best AI image generator model?
-
OpenAI DALLE-2 Outpainting Fills Image Blanks
By
–
DALLE – 2 outpainting. @OpenAI fills in the blank around and between images, blending styles and mediums. pic.twitter.com/A7RtiuOfwS
— AI Breakfast (@AiBreakfast) 26 novembre 2022DALLE – 2 outpainting. @OpenAI fills in the blank around and between images, blending styles and mediums.
-
Capabilities and use cases for GPT-3
By
–
So what can you use it for? GPT-3 will attempt to answer virtually any text prompt you give it. It can answer questions, write stories, poems, create lists, write research papers, and it can write code. A lot of code. The best way to describe what it can do is to just try it.
-
How Language Models Function as Probabilistic Text Predictors
By
–
When a user provides the text input, the system analyzes the language and uses a text predictor to create the most likely* output. *Does not necessarily mean the most factually accurate output, but rather the most "likely" output based on the data from the training library.
-
GPT-3 Training Data Composition Breakdown
By
–
GPT-3 has been trained on 45 TB of text data from different categories: ⬩Common Crawl (8 years of raw web page crawling) ⬩WebText (The text of Reddit posts with 3+ upvotes) ⬩Books (The internet-based books corpora) ⬩Wikipedia Data is then "weighed" as such:
-
How AI Models Transform Input into Predictions
By
–
This means that it can take user's input text and transform it into what it predicts the most useful result will be based on the patterns recognized from the parameter library.