.png)
In short, Artificial Intelligence or AI is using technology to predict outcomes or patterns. AI takes “unstructured” or random data and uses “cognitive technology,” such as computer vision, machine learning, natural language processing, speech recognition, and robotics, to think like a person (a really smart person). The data, which ranges from images, videos, sounds, text, or quantities, are put into certain categories for processing, similar to the human brain. As the data are categorized, AI can understand it in real time, with real application.
Whiles specific uses of AI include predictive medicine, identifying financial trends, and misinformation campaigns, it is divided into 7 general categories:
1. Hyperpersonalization: Mimics an individual person.
This pattern created by AI uses machine learning to profile an individual, focusing on what makes him/her unique. The profile adapts over time, learning from the host, and is used in marketing, advertising, and even credit assessment.
In marketing and advertising, AI looks to target advertisements based on individual preferences, needs, and budgets by displaying relevant content, products, and knowing how the individual values influences. For example, would he/she prefer a product with a 4.2-star rating based on 3,500 reviews or a 4.6-star rating with only 6 reviews? This pattern desires to transform behavioral profiling from subjective to objective science.
2. Autonomous systems: Remove the need for a person.
Autonomous systems look to accomplish a task or interact with surroundings without (or with minimal) human involvement. This is applied both to physical, hardware autonomous systems as well as software or virtual autonomous systems (software “bots”). Examples are robots, self-driving cars, unmanned aerial vehicles (UAVs), and automatic document generators.
The Society for Automotive Engineers has defined six levels of autonomous systems, ranging from Level 0 to Level 5. An example of Level 0 autonomy is an autonomous vehicle requiring full human control of all operation and decision making, while a Level 5 autonomous vehicles does not require human attention, lacks brakes, steering, and can go anywhere it pleases.
3. Predictive analytics and decisions: find decisions with more reliable outcomes.
Using patterns and machines to make better decisions and predict future outcomes based on data, previous interactions, and learned human behavior. Often used to augment human analyses, these models predict failures, values, pricing, and trends to determine potential problems, identify better investments, and augment existing human intelligence.
4. Conversational/human interactions: mimic human conversation.
The conversational / human interaction pattern is machines interacting with humans through voice, text, images, and writing to boost interactions between machines and humans, as well as between humans and other humans. This does not include machine to machine communication because machines do not use human forms of communication. The purpose is to allow machines to communicate with people as well as people do with each other through virtual chats, voice assistants, content creation for video and audio. AI focused on conversational / human interactions can analyze moods, sentiments, and intent from human speech.
5. Patterns and anomalies: find what doesn’t fit in the pattern.
Reflecting the name, the purpose of this use of AI is to determine which data belongs in a pattern and which data does not. Primarily used to detect fraud, malice, errors, and used with intelligence monitoring, AI looks for predictive data to identify the abnormality. This is a cognitive function for which machines are better equipped than humans.
6. Recognition systems: recognize words, sounds, and images.
This pattern uses machine learning and other cognitive approaches to identify and determine objects or other desired things to be identified within some form of unstructured content, such as images, video, audio, or text that can later be labeled and tagged. The machine is essentially able to identify, understand, and sort the content. Examples are facial recognition software, handwriting and text recognition, sound, and audio recognition (“Press 1 for operator”). This is the most commonly used type of AI.
7. Goal-driven systems: trial and error for improvement.
Using machines to try new approaches to problems using trial and error in pursuit of better solutions. Examples are scenario simulation (flight simulators), game playing, resource optimization, and iterative problem solving.
Many of the different uses of AI can be combined to achieve more advanced applications provided that the user understands the purpose of each type.
Unlike old-fashioned computer programming in which a human created the rules and boundaries in which the machine operates, AI allows the computer to learn and develop new skills from the information presented to it over time. While many find such possibilities frightening, especially as only 1 U.S. Law regulates its use, AI is already proving helpful in medical diagnoses, inventory management, and weather forecasts.