Generative AI

Adults & Seniors

by LearnX Digital

Generative AI

  • Share

Generative AI

 

Flex your creative muscles with Generative AI!

From generating symphonies to literature, these machines can turbo-charge your creative prowess with a few taps of the finger. But that's not all—they're also transforming industries and sparking new waves of innovation.

Read on to discover how it's empowering us to achieve new heights in innovation!



What is Generative AI?

 

Generative AI, short for Generative Artificial Intelligence, refers to a class of artificial intelligence systems designed to generate new content, or output that mimic or resemble human-created content. Unlike traditional AI systems that rely on explicitly programmed rules (e.g. distinguish a car from a bus), generative AI leverages machine learning techniques, particularly within the domain of deep learning. Deep learning refers to teaching a computer to think and make decisions on its own by showing it a large quantity of examples.

 

 

How do Generative AI tools like Large Language Models work?

 

Large language models, such as GPT-3 (Generative Pre-trained Transformer 3), are sophisticated artificial intelligence models designed for natural language understanding and generation tasks. They can understand, generate, and respond to human language in a conversational way. 

Here is an analogy of how Large Language Models work:

Suppose you are training a friendly new alien who has just come to Earth. This alien has a humongous brain with a basic understanding of communication and loves math but doesn’t know any English at all.

How can it learn to tell stories?

 

Alienpic

(Generated with AI; Bing Image Generator)

 

You bring the alien to a magical library containing all the books in the world. It reads everything, but it doesn’t just read words one at a time or memorise words. The alien reads in a smarter way, identifying patterns and relationships between words. To learn even better, it practices telling stories.

It takes a story and tries to predict the next word, for example:

Once upon a ___,

Time (70%)

Twice (20%)

Lake (10%)

If it’s wrong, it readjusts its brain connections and tries again.

Now, the alien has a basic understanding of patterns in English (one might say it is pre-trained). But it’s only good at predicting the next word, rambling on and on.

You then sit the alien down and have a long chat. You ask it for many stories based on different prompts and genres and give the stories a score. By giving the alien feedback, it gradually learns which types of responses you prefer.

Through this training with the alien, it learns to tell intriguing stories.

 

 

 

Did you know?

 

Large Language Models like ChatGPT are not new. You might have used tools based on such technologies, like Google Translate or autocorrect.

 

A vast amount of data and computing power made tools like ChatGPT incredibly powerful beyond its predecessors!

 

 

 

 

Curious to see how Generative AI is used in our libraries?

 


Get to know Generative AI through StoryGen, an ExperienceIT prototype, which lets you add a twist to well-loved stories and see them come alive through an immersive visual experience.

StoryGen is available at the Stories Come Alive Room at Level 1 of Punggol Regional Library on weekdays, when there are no programmes, until 30 October 2024.

 

 

  • How is Generative AI applied in various industries today?

  • Ethical Considerations about Generative AI

  • How can I use Generative AI Responsibly?

  • Where can I learn more?

  • Try it yourself!