Artificial Intelligence Chapter 7A Flashcards
When was AI first coined?
AI was first coined in 1950 alongside the LISP programming language
When was the first AI winter?
The first AI winter took place in 1973
What caused the sudden appreciation of AI again?
1) Powerful intensive computers became widely available albeit at a higher cost (GPU)
2) Big Data: The availability of big data due to smartphones and the internet of things accelerated the adoption of AI
3) Software algorithms: Deep Learning, Machine Learning
What is the difference between supervised and unsupervised learning?
- Supervised learning required labelling of data (regression & classification)
- Unsupervised learning DOES NOT require the labelling of data (clustering)
What is the concept of reinforcement learning?
Reinforcement learning makes use of “agents” to help learn and maximize the software by giving “rewards” in that environment.
use of ‘agent’ that learns to maximise ‘reward’ in an environment
What are the several Deep Learning Methods? What are each of them used for?
- Artificial Neural Network (Used in the processing of large number-type data)
- Convolution Neural Network ( Used in the recognition of images)
- Recurrent Neural Network (Used in the recognition of audio)
- Deep reinforcement learning
- Transfer Learning (Used to repurpose the AI to target and transfer its algorithms to another use case)
What is a hidden layer?
A hidden layer consists of multiple learnable parameters, it is essentially an algorithm that can learn & improve by itself.
What is a deep neural network?
Deep neural network are multiple layers of hidden layers. They allow for much more sophisticated algorithms to be learnt
What are the concerns about AI?
- Job Loss
- Misuse of AI (Deepfakes)
- AI explainability
- AI bias (Who programmed the AI)
- AI in ethics (Weaponizing AI)
