AI Terms Authors Need to Know

/

Artificial Intelligence (AI) is exciting and terrifying at the same time. Regardless of how you feel about it, it’s here, and it’s popping up everywhere. Here are some terms you need to know. I’ve included some great resources at the end that provide a lot more detail and examples.

Agentic AI: Systems designed to be autonomous. They can make decisions and do actions with no or limited human interactions.

AI Ethics: A multidisciplinary field that studies how to optimize the beneficial impact of AI while reducing risks and negative results.

Algorithm: A rule that a machine follows to complete a task.

Artificial Intelligence/AI: This is where machines act in a way that mimics human intelligence. Sometimes, it’s called cognitive computing.

Autonomous: A machine that can perform tasks without human intervention.

Bias/Machine Learning Bias: Slanted results caused by human bias, training data, or algorithms that lead to distorted results and possible negative outcomes.

Big Data: These are source or sets of data that are too large to process by traditional computers or data processing systems.

Chatbot: A process created for computers to communicate or interact with humans. They mimic human conversations.

Cognitive Computing: A model that mimics human thought processes.

Deep Fake: AI-generated images, videos, or audio to perpetrate a hoax or spread fake information.

Deep Learning: This is a function of AI that mimics human learning. It learns how data is structured rather than by the algorithm or rules that program it to do one task.

Emergent Behavior/Emergence: When an AI program shows unpredicted or unintended capabilities.

EU AI Act: A European Union regulation for responsible AI deployment and privacy.

Generative AI: This is where a system can process data and generate something new instead of regurgitating data that already exists. The system can create text, videos, code, and images.

Guardrails/Controls: These are frameworks or checks to ensure AI systems operate ethically and legally.

Hallucinations/Fabrications: The systems cannot discern truth from fiction from what it is trained on and what it generates.

Large Language Model/LLM: A training method of AI where huge amounts of data are used to teach the understanding of natural language.

Machine Learning: Part of AI learning that centers on algorithms that can “learn” the patterns of training data and make recommendations about new data. This pattern recognition ability enables machine learning models to make decisions or predictions without programmed instructions.

Natural Language Processing/NLP: This is a method where the computer takes structured data and turns it into text or speech.

Prompt: The way humans interact with a system. It in an instruction for a task.

Turing Test: Named for mathematician Alan Turing, this test checks the machine’s ability to pass for a human.

Vibe Coding: A method for developing software where the user guides an AI system through natural language prompts to create, refine, and test code. This is an alternative to writing the code. The term was coined by Andrej Karpathy.

Voice Recognition/Speech Recognition: A way for humans and computers to interact. The computer listens and interprets human speech.