Types of AI
This type of AI has a limited function or can only perform a specific task, like Netflix recommendations or Siri.
What is Weak or Narrow AI?
This simplest form of machine learning uses algorithms that learn by example from training data consisting of inputs paired with correct outputs.
What is Supervised Learning?
Systems with interconnected nodes that work similarly to human brain neurons, processing information through layers.
What are Artificial Neural Networks?
A period when funding and interest in AI research significantly decrease, with the first one occurring between 1974 and 1980.
What is an AI Winter?
A significant AI limitation where inaccuracies or prejudices in the training data lead to skewed or unfair results.
What is Bias?
A computer system that acts like a human expert in a specific subject area, often used to advise non-experts
What is an Expert System?
This type of machine learning uses algorithms trained by rewarding positive behavior and penalizing incorrect actions, adapting to its environment to maximize rewards.
What is Reinforcement Learning?
This layer in a neural network receives the initial data input.
What is the Input Layer?
Popularized by Vernor Vinge and Ray Kurzweil, this term describes the hypothetical future where AI growth becomes uncontrollable and surpasses human intelligence
What is the Singularity?
The principle that AI systems should be designed so humans can easily understand their decision-making processes to build trust.
What is Transparency?
This test, developed by Alan Turing, determines if a machine exhibits intelligent behavior indistinguishable from a human, famously passed by a chatbot posing as a 13-year-old boy.
What is the Turing Test?
Used to find underlying patterns in unlabeled data sets, often employing cluster analysis to group data.
What is Unsupervised Learning?
Processing in a neural network occurs here through a system of weighted connections between the input and output layers.
What are Hidden Layers?
Coined by roboticist Ken Goldberg, this concept envisions a future where AI and robots work alongside humans rather than replacing them entirely.
What is Multiplicity?
The increased use of technology in a process, like in manufacturing or food preparation, which reduces the need for human involvement and displaces jobs.
What is Automation?
The hypothetical future point when AI becomes so advanced it surpasses human intelligence.
What is Super AI or the Singularity?
Machine learning algorithms used to recognize either observed physical patterns or mathematical patterns, applied in areas like data analytics and image recognition.
What is Pattern Recognition?
A type of machine learning used for complex problems that employs artificial neural networks with more than three layers.
What is Deep Learning?
A pioneer in this field, which studies communication and control in living things and machines, alongside John McCarthy and Marvin Minsky, researched AI starting in the 1950s.
Who is Norbert Wiener (and what is Cybernetics)?
This component of an expert system uses programmed rules (like if/then statements) to interpret facts in the knowledge base.
What is the Inference Engine?
AI that can develop consciousness, self-awareness, emotions, and possess full human cognitive abilities
What is Strong AI (also known as Full AI or AGI)?
This machine learning application teaches computers to understand human language, used in text analysis, translation, and chatbots.
What is Natural Language Processing (NLP)?
The ability of neural networks to learn and model these types of complex, real-world relationships is a key advantage.
What are Non-linear relationships?
Developed in 1965 at Stanford, this was the first expert system, designed to analyze chemical compounds.
What is Dendral?
A modern-day Turing Test used on websites to distinguish human users from automated bots.
What is CAPTCHA?