Self-driving cars

Cryptocurrencies

Mobile payment using NFC

Tours in Virtual Reality

Large Language Models

Chatbots

Generative AI models

Computer Vision

Large Language Models

Chatbots

Neural networks

Generative AI

AI does not require any code at all and can independently program itself.

Regular programs only operate within defined scenarios while AI can learn by itself

Regular programs are the only way to build chatbots

AI is only possible with special,  very expensive computers

A neural network that solves computer vision tasks

A deep learning  model for chatbots which converse like humans

A machine learning algorithm to learn patterns in numerical data

A generative AI model which can take an input text and produce an output image

For an input sentence, output if it true or false

For an input word, output a similar word

For an input sentence, output a human-like discussion about it

For an input sentence, output the most likely next word 

Large Language Models

Chatbots

ChatGPT

Generative AI

From input to output, the layers gradually move from simpler to more complex forms

Each layer focuses on one aspect: colour, edges, shapes etc.

Nobody really knows what happens inside the neural network but it works!

From input to output, the layers gradually move from complex to simpler forms

An AI model that learns from data without human interference 

An AI model which learns from input data without being told what is the correct output

A machine learning algorithm which needs no training data 

An AI model which can find its own data for training, without human input

It is a set of algorithms which learn to reply like a human by being trained with huge amounts of text

All machine learning models are generative AI

All machine learning models are neural networks

It is a set of algorithms which learn statistical models by being trained with data