What is GenAI
Just one month ago, on May 10, 2023, Google CEO Sundar Pichai provided a glimpse of Google Genrative AI while kicking off Google I/O 2023. AI is a broad term often used to describe all sorts of advanced computer systems. Most of what we see in AI today is really machine learning — endowing computer systems with the ability to learn from examples.
How Traditional Machine Learning Works?
We call machines programmed to learn from examples “neural networks.” The research into biological neurons led to the creation of a new computing paradigm, the artificial neural network. With artificial neural networks, we mimic the behaviour of biological neurons with simple mathematics. It creates an adaptive system that computers use to learn from their mistakes and continuously improve. Thus, artificial neural networks attempt to solve complicated problems, like summarising documents or recognising faces, with greater accuracy. You can use the following simple neural network, which features a single neuron (aka perceptron)for understanding. Perceptrons are a type of artificial neuron that takes several binary inputs like x1, x2,… Xm and produces a single binary output.
x1 ,x2 and Xm are the input values, and w1 and w2 are weights that represent the strength of each connection to the neuron. Y is the output given by model. Tensorflow’s playground shows you how neural networks work and affect their output.
One main way they learn is by being given lots of examples to learn from, like being told what’s in an image — we call this classification. If we want to teach a network how to recognise a car and bike, that would involve a human introducing the network to lots of examples of what a car and bike look like and tagging (labelling the datasets) those photos accordingly. That’s how the model learns to distinguish between cars, bikes, and other details in an image.
How do Large language models work?
Large Language Model (LLM): A large, general-purpose language model can be pre-trained and then fine-tuned for specific purposes. Large language models are trained to solve common language problems. Like
- Text Classification
- Question Answering
- Document Summarization
- Text Generation
Also used for retail, commerce, and entertainment areas after fine-tuning on LLM.
Large: large training datasets, large number of parameters
General Purpose: Commanality of Human Languages
Pre-trained and fine-tuned: pre-trained on large datasets; fine-tuning with specific aims with much smaller datasets.
In traditional training models, we have hard-coded rules. Suppose we are building a model that will classify whether a cat is a cat or not.
In neural networks, we train models with datasets and their labels, e.g., this is cat for images containing cats, and this is not cat for images containing other animals.
In a generative large learning model (LLM) like Lambda, GPT, or PaLM, we are supposed to give one book related to cats, and the model will learn this data and provide you with output.
Language models are already out there helping people — you see them show up with Smart Compose and Smart Reply in Gmail, for instance. And language models like Chatgpt and Google’s Bard.
What About Generative AI
Generative AI is a type of Artificial Intelligence that creates new content based on what it has learned from existing content. GenAI produces various type of content images,text,audio, synthetic data. Genrative AI is a subset of deep learning. Deep learning uses artificial Neural Networks allowing them to process more complex pattern than Machine learning.
There are two types Deep Learning Model Discriminative and Genrative. Discriminative technique trained on labeled data and use for classification or prediction where Generative technique generate new data i.e. similar to the data it was trained on.
Types of Generative AI based on data shown below.
A generative model can take what it has learned from the examples it’s been shown and create something entirely new based on that information. Hence the word “generative!” Large language models (LLMs) are one type of generative AI since they generate novel combinations of text in the form of natural-sounding language. And we can even build language models to generate other types of outputs, such as new images, audio and even video, like with Imagen, AudioLM and Phenaki.
GenAI Studio
Generative AI studio lets you quickly exploreand customize genAI model that you can leverage in your applications on google cloud e.g. Building chatbots, semantic analysis etc.
Gen AI studio helps developers to create and deploy GenAI models
- Fine-tune Model
- Deploy Model to production
- Create Chatbot
- Image generation
- Semantic Analyis
Generative AI App Builder
Generative AI app builder creates generative AI Apps without having to write any code. Gen AI App Builder have drag and drop interface that makes it easy to design and build app. Visual Editor helps to customize and edit app content. There is inbuilt search engine allows user to search for information within app. It has conversational AI Engine that help user to interact with the app using natural langauge.
Useful Links:
