AI terminology

Hi, in this post I’m going to insert general terminology that is used in IA (Artificial Intelligence).

What is AI?

Artificial Intelligence refers to the simulation of human intelligence in machines that are programmed to think and learn like humans.

Is there more subfields of AI?

Yes it is: machine learning, natural language processing, computer vision, robotics, and more.

Which task could do AI?

AI could understanding and interpreting complex data, making decisions, recognizing patterns, and even mimicking human speech and behavior.

How AI can do those tasks?

Using algorithms and models that enable them to process large amounts of data, extract meaningful insights, and make predictions or take actions based on the information they have learned.

Which are the 2 main types of AI?

Narrow AI (systems that are designed for specific tasks and have a limited scope of capabilities) and General AI (develop machines that possess the same level of intelligence and understanding as humans.)

Some real products that are using AI

Recommendation Systems: Platforms like Netflix, Amazon, and Spotify

Autonomous Vehicles: Tesla, Waymo, AI algorithms for perception, decision-making, and navigation.

Image and Object Recognition: Google Photos, Facebook (AI automatically identify and tag people, objects, and locations in photos)

Chatbots: chatbots (chat) to interact with users, answer questions, and provide support, implemented by many sites.

Visual accessibility: for example Be My Eyes company implement visual assistance for instantaneous image-to-text generation.

What is the difference between Model and algorithm in AI?

Algorithm: is a step-by-step procedure followed to solve a specific problem, in AI represents the underlying logic or methodology used to train a model or make predictions (is a mathematical or computational formula that dictates how data is processed and transformed.). Examples of AI algorithms include decision trees, support vector machines (SVM), k-means clustering, and neural networks.

Model: In AI, a model refers to a representation or approximation of a real-world phenomenon, system, or process. It is built using algorithms and trained on data to learn patterns, make predictions, or perform specific tasks. A model is the result of applying an algorithm to data during the training process. It captures the learned information and can be used to make predictions or generate outputs for new, unseen data.

What is training phase in AI?

Specific stage in which a machine learning model learns from a given dataset (check this link you will see a dataset of bean images) to improve its performance on a particular task.

How many models are there in AI?

AI models is difficult to determine precisely because new models are constantly being developed and existing models are being improved upon.

Some commonly used and influential AI models:

Linear Regression, Logistic Regression, Decision Trees, Random Forest, Support Vector Machines (SVM), Naive Bayes, Artificial Neural Networks, Transformer Models (language translation and text generation examples include BERT (Bidirectional Encoder Representations from Transformers) and GPT (Generative Pre-trained Transformer) models. Reinforcement Learning Models, Generative Adversarial Networks (GANs).

What is generative AI?

Category of artificial intelligence techniques and models that are designed to generate new content or data that is similar to, or inspired by, existing examples.

Generative AI models operate by learning patterns and structures from training data and then generating new data based on those learned patterns.

Generative AI has numerous applications across different domains. For example: it can be used in creating realistic deepfake videos, generating new artwork or music, text synthesis, data augmentation and more.

What is chatGPT?

ChatGPT is an AI language model developed by OpenAI

Which model is using ChatGPT?

Is built using OpenAI’s GPT (Generative Pre-trained Transformer) architecture.

What is OpenAI?

Is an artificial intelligence research organization and company. Founded in December 2015 as a nonprofit organization and later transformed into a for-profit company named OpenAI LP in 2019.

Where can I find some free AI Models?

TensorFlow Hub is a platform that hosts a variety of pre-trained models across different domains, including computer vision, natural language processing, and more.

You can play with this model where you can upload a food image from your computer and then the model will tell you what is your image about.

For example: I uploaded a taco image and the Food classification model gave me an approximation result of 0.956 which is almost perfect.

OpenAI GPT Models OpenAI has released several versions of the GPT language model.

PyTorch Models These models can be loaded and used for tasks like image classification, object detection, text generation, and more.

Hugging Face Transformers: provides a wide range of pre-trained models for natural language processing (NLP) tasks. The models are available through the Transformers library, which includes architectures like BERT, GPT, RoBERTa, and many others. These models can be used for tasks such as text classification, named entity recognition, machine translation, and question answering.

By Cristina Rojas.