IBM: Introduction to Artificial Intelligence

Artificial intelligence (AI)

refers to the ability of a machine to learn patterns and make predictions; a field that combines computer science and robust datasets to enable problem-solving.

Augmented vs Artificial Intelligence

Augmented intelligence has a modest goal of helping humans with tasks that are not practical to do. In contrast, artificial intelligence has a lofty goal of mimicking human thinking and processes

Analysis

AI services can take in (or “ingest”) enormous amounts of data. They can apply mathematical calculations to analyze data, sorting and organizing it in ways that would have been considered impossible only a few years ago.

Predictions

AI services can use their data analysis to make predictions. They can, in effect, say, “Based on this information, a certain thing will probably happen.”

Narrow AI

is designed to perform specific tasks rather than general intelligence. Unlike human cognitive abilities,

Broad AI

refers to artificial intelligence that possesses general cognitive capabilities, enabling it to understand, learn, and perform any intellectual task that a human can do.

General AI

machines that can perform any intellectual task that a human can.

artificial superintelligence (ASI)

is a form of artificial intelligence that surpasses human intelligence across all fields, enabling it to outperform humans in virtually every cognitive task.

tabulation

“slicing and dicing” data to give it a structure, so that people can uncover patterns of useful information

Era of Tabulation

a time when machines helped humans sort data into structures to reveal its secrets

Era of Programming

a period characterized by the development of software that enabled automated data processing and problem-solving, laying the foundation for modern computing and AI.

Turing Machines

machines capable of simulating any algorithmic process, fundamental to understanding computation and algorithms in computer science.

Dartmouth Conference

A pivotal event in 1956 where researchers gathered to discuss the potential of artificial intelligence, marking the birth of AI as a recognized field.

1st AI Winter

caused by high expectations from end users and reduced funding.This period, occurring in the 1970s, saw a decline in AI research due to unmet promises and disillusionment among researchers and investors.

2nd AI Winter

caused by unmet expectations and computing power.This period in the late 1980s to early 1990s saw a decline in AI research funding and interest due to the inability of AI systems to deliver on promises.

Deep Blue

was a computer chess-playing system developed by IBM that became famous for defeating world champion Garry Kasparov in 1997, demonstrating significant advancements in AI and computational power.

Watson

is an IBM AI system known for its natural language processing capabilities. It gained fame by winning the quiz show Jeopardy! against human champions in 2011, showcasing the potential of AI in understanding and processing human language.

AlphaGo

is an AI program developed by DeepMind that plays the board game Go. It became widely known for defeating world champion Lee Sedol in 2016, illustrating the advanced capabilities of machine learning and neural networks in complex game scenarios.

Project Debater

the first AI system capable of engaging with humans on complex topics in a live debate.

Data

raw information that might be facts, statistics, opinions, or any kind of content that is recorded in some format.

Structured Data

typically categorized as quantitative data and is highly organized; information that can be organized in rows and columns

Unstructured/Dark Data

typically categorized as qualitative data. It cannot be processed and analyzed by conventional data tools and methods; lacks any built-in organization, or structure. Makes up 80% of the world’s data today

Semi-structured data

is the “bridge” between structured and unstructured data. It doesn't have a predefined data model. It combines features of both structured data and unstructured data

Meta data

used by semi-structured data to identify specific data characteristics and scale data into records and preset fields

Advantages of Machine Learning Process

• It doesn’t need a database of all the possible routes from one place to another. It just needs to know where places are on the map.

• It can respond to traffic problems quickly because it doesn’t need to store alternative routes for every possible traffic situation. It notes where slowdowns are and finds a way around them through trial and error.

• It can work very quickly. While trying single turns one at a time, it can work through millions of tiny calculations.

• It can also predict and learn

Deterministic/Classical Learning System

Must include an enormous, predetermined structure of routes—a gigantic database of possibilities from which the machine can make its choice; the essence of a computer program

Probabilistic/ Machine Learning System

A system that learns from data patterns and experiences instead of relying on a fixed set of rules or pre-defined outputs, allowing it to make decisions based on probabilities. It continuously updates its models to improve accuracy and performance over time.

Supervised Learning

providing AI with enough examples to make accurate predictions. Requires telling the model:

1. What the key characteristics of a thing are, also called features

2. What the thing actually is

Labeled Data

providing AI with enough examples to make accurate predictions; required by all supervised learning algorithms

Classification Problem

A type of supervised learning task where the goal is to predict the categorical label of input data based on training examples.

Unsupervised Learning

A type of machine learning where the model is trained on unlabeled data, discovering patterns and structures without predefined categories.

Reinforcement Learning

A type of machine learning where an agent learns to make decisions by receiving rewards or penalties based on its actions within an environment.

Environment

The setting in which an agent operates in reinforcement learning; it includes the state space, action space, and reward structure that influence the agent's learning and decisions.

Agent

refers to the entity that takes actions within an environment to achieve a goal, learning from the consequences of its actions in order to maximize cumulative rewards.

State

A situation returned by the environment after each action taken by the agent

Action

the moves taken by the agent within the environment

corpus

a collection of written or spoken material used for linguistic analysis or training machine learning models.

reward

the positive numerical feedback received when the comparison of the prediction to a agent’s corpus is positive

penalty

the negative numerical feedback received when the comparison of the prediction to a agent’s corpus is negative

Instrumented Planet

a framework or system for gathering data from various sources on a planetary scale, often used in AI modeling and analysis.

AI Everywhere

the idea of how AI in the future will move into all industries