<aside>
<img src="/icons/flag_red.svg" alt="/icons/flag_red.svg" width="40px" /> It’s important to note that like any other tool, the use of Generative AI comes with ethical considerations, the most addressed one these days being the use of copyrighted content. Since generative AI tools started being a part of our lives with exponential acceleration, it's important to use them responsibly, considering factors like user privacy, data security, intellectual property rights, and the potential for misuse. Please don’t forget to check out Principles, Dos & Don’ts to learn more about how to use AI more responsibly.
</aside>
Generative AI is a subset of artificial intelligence that focuses on creating new content based on a variety of inputs such as texts, sounds, images, or other media.
Generative AI models are trained on a large amount of data. They learn the patterns and structure of that training data and then use that to generate new, original content that has similar characteristics. Essentially, it's a way for AI to create something new instead of just classifying or interpreting data.
<aside> <img src="/icons/yin-yang_blue.svg" alt="/icons/yin-yang_blue.svg" width="40px" /> Generative AI models use a specific type of Neural Network called a Generative Adversarial Network to identify different data patterns and structures.
GAN consists of two parts: a generator network that creates new data, and a discriminator network that evaluates the data and tries to tell the difference between real data and the data created by the generator. (It’s basically a love-hate relationship where they try to beat each other but also get stronger thanks to this competition.)
</aside>
By the type of content they generate:Text Generation
These AI models can write essays, create poetry, answer questions, draft emails, and much more. They're trained on a diverse range of internet text, and they work by learning the statistical structure of the text they are trained on and then generating new text that follows this learned structure.
Example Tools:
well…
Image Generation
AI image generators are designed to create new images based on a set of input parameters or conditions. These systems use machine learning algorithms that can learn from large datasets of images, allowing them to generate new images that are similar in style and content to the original dataset. They can also create artwork or alter image attributes.
Example Tools:
the possibilities are endless
Audio Generation:
These AI tools can generate audio content, including music, human speech, and ambient sounds. Audio generation involves understanding and replicating complex patterns found in audio data.
Example Tools:Video Generation:
Video generation can create every video content from short animated clips to full-length films. It involves synthesising visual frames and potentially audio to generate sequences that mimic the style and content of the training data.
Example Tools:https://www.youtube.com/watch?v=KrjL_TSOFrI
https://www.youtube.com/watch?v=d-8DT5Q8kzI
By the type of conversion tasks:Text-to-Speech (TTS)
This involves converting written text into spoken words. It's widely used for accessibility purposes, voice assistants, and more. It works by training a model on a large dataset of spoken language, allowing it to learn how to produce speech sounds corresponding to text input.
Example Tools: