AI

Question 1

What is XAI (Explainable AI)

Answer 1

XAI are models that aim to be as well performing as they are interpretable. XAI is designed so that a human user can understand how it reached a particular decision/outcome.

Question 2

Why is XAI important?

Answer 2

XAI is important as it provides trust and transparency. This is important in high stake areas such as medicine. Furthermore, it allows for adjustments and improvements, because if we know how it works we can identify the problems and solutions.

Question 3

What is interpretability in AI?

Answer 3

Interpretability refers to the level of understanding of HOW an outcome is reached.

Question 4

What is explainability in AI?

Answer 4

The level of explanation there is FOR the outcome of a model. (Also acts as an accuracy proxy of the model)

Question 5

What is fidelity in AI?

Answer 5

Fidelity refers to how WELL the explanation matches the actual behaviour of the model.

Question 6

What is comprehensibility?

Answer 6

The EASE of which users can understand a provided explanation.

Question 7

What are the four types of evaluations humans perform to ensure user ability of XAI?

Answer 7

Binary forced choice, forward simulation/prediction, counterfactual simulation, surveys

Question 8

What are the two approaches for XAI and what is their focus?

Answer 8

Intrinsic: more interpretable
Post-hoc: better explanations

Question 9

What are examples of intrinsic models? and why is this?

Answer 9

Linar models, decision trees, k-Nearest neighbours, rule-based systems, personalised interpretability systems. These are more interpretable models to human users as long as they are relatively small and less complex. We can understand how they reached the outcomes they did by looking at them.

Question 10

What can we use to reduce the number of features in models? Why do we want to minimise weights?

Answer 10

Lasso regression. Implicity needs as many weights to be zero as possible for increased understanding.

Question 11

What does k-Nearest neighbours do?

Answer 11

uses existing instances to explain new instances

Question 12

What are some approaches to maximise intrinsic interpretability in neural networks?

Answer 12

add interpretability constraints, visualise network attentions, employ attention matrix, forcing weights and parameters to zero, model-specific complexity measures, consider meaning of intermediate outputs

Question 13

What is the con of more localised sampling and the con of generalised sampling?

Answer 13

More localised sampling might not capture enough information about a complex model.
More generalised sampling might be too complex to capture a linear model.

Question 14

What are the two types of AI systems social media incorporates?

Answer 14

Recommender systems and harmful content classifiers

Question 15

What are recommender systems?

Answer 15

Recommender systems are what push content at you. They curate your feed using a positive feedback loop.

Question 16

What are harmful content classifiers?

Answer 16

Harmful content classifiers withhold harmful content from users by removing or downlinking risk content. This content moderation works using supervised learning and uses a harmful content rating.

Question 17

What types of AI are “coming” for human jobs?

Answer 17

Generative AI is the leading impact - LLMS. Gig platforms such as uber and door dash are too. This is because they are replacing human organisation/management. AI is able to effectively sort and compare information, enabling it to be a helpful tool in the recruitment process too. It can sort through CVs and judge applicant responses in AI interviewing systems.

Question 18

What is a concern for the AI recruitment process?

Answer 18

Bias. There is an ethical concern that AI may perpetuate bias in its recruitment process due to misrepresentation in its data, causing it to overlook some individuals.

Question 19

What are the three possible scenarios for the future of AI and jobs?

Answer 19

“Enabling AI”
- benefiting humans, causing us to be more productive and work more effectively. no loss to jobs, just less workload
“Replacing onshore”
workers are displaced into lower value work. but we can regain profit from making AI jobs NZ owned and taxed.
“Replacing offshore”
again displacing workers into lower value work but AI systems are owned offshore therefore we cannot regain profit

Question 20

What is a solution for reducing further inequality with the rise of AI in jobs?

Answer 20

tech taxes.

Question 21

What laws have been set up across the world to ensure safety around AI?

Answer 21

EU’s AI Act (UK)
Bletchey Park AI Safety Summit
Executive Order on AI (US)
Interim Measures on Gen AI (China)

Question 22

What is the Maori Algorithmic Sovereignty?

Answer 22

MASov acknowledges that any data that is about Maori, by Maori, or about Maori environment should be subject to Maori governance. They recognise that if data under/misrepresents Maori, the models will perpetuate bias and stereotypes that can further worsen outcomes for Maori.Maori should have control over how their data is collected, stored and used.

Question 23

Explain the principle of Whakapapa in AI?

Answer 23

Transparency: there should be transparency of the algorithm - who is involved, the motives, deployment, explainability etc. Maori participants should be given transparency on their data.
Data relationship: how is Maori data used throughout the algorithm. This aligns with the MASov.
Sustainability: should ensure that the algorithms output benefits Maori for the long term and that there is a sustainable positive outcome.

Question 24

What is ‘one hot’ encoding?

Answer 24

encoding for words by which there is one unit for each word

Question 25

What does implementing an attention mechanism in a neural network do?

Answer 25

It allows the model better access to previous context than a recurrent network. This leads to learning which words in the context are the most important for predicting the next word.

Question 26

What did OpenAI use to remove harmful content from GPT-4 before it was trained.

Answer 26

Text classifiers

Question 27

What should you do if you are building a decision tree and at a particular node, all the data belongs to a single class?

Answer 27

You should mark this node as a leaf node and assign it the class label.

Question 28

Is clustering an example of a reinforcement learning task?

Answer 28

No, clustering is an example of unsupervised learning as it learns to group similar data points together based on their features without any prior labelling or guidance.

Question 29

In regression tasks, is the output continuous?

Answer 29

Yes, the data is continuous numerical- e.g, house prices, can be measured down to cents and infinite prices.

Question 30

Describe artificial intelligence

Answer 30

Artificial Intelligence is a broad field of computer science that focuses on creating intelligent agents that can reason, learn, and act autonomously. AI aims to simulate human intelligence and problem-solving capabilities in machines. There is narrow AI, general AI and super intelligence. Some AI are also transformers, which means they are trained broadly on things and fine-tuned to specify in something - GPT-4 is a transformer model. Self driving cars such as Tesla are applications of AI too.

Question 31

Describe machine learning

Answer 31

Machine learning is a subset of artificial intelligence that focuses on developing algorithms and models that allow computers to learn from data and improve their performance on a specific task without being explicitly programmed. In essence, machine learning enables computers to "learn" from experience. An example of machine learning is recommendation systems such as the recommender systems seen in social media that push peoples preferences at them.

Question 32

Describe deep learning

Answer 32

Deep learning is a subset of machine learning that uses neural networks with multiple layers to learn complex patterns from data. AI Image generation is an example of a deep learning model.

Question 33

Describe the relationship between artificial intelligence, machine learning and deep learning

Answer 33

Artificial intelligence is a field of computer science -> machine learning is a subset of AI -> deep learning is a subset of machine learning. AI is the general field of creating intelligent agents. ML is a specific approach within AI that focuses on learning from data. DL is a specialized technique within ML that uses deep neural networks.

Question 34

What is overfitting in the context of supervised learning and why does it occur?

Answer 34

Overfitting happens when a model fits too closely and memorises the patterns of the training data, causing it to be unable to generalise to unseen data.

Question 35

Explain how to reduce the chance that a decision tree overfits

Answer 35

To avoid a decision tree from overfitting, ensure the model is not too complex and doesn't have too much 'noise'. Further, train the model on test data, so it doesn't perfect training data patterns and is able to generalise to unseen data. Prune: set a maximum depth or minimum number of samples per leaf to limit the tree's growth. Remove irrelevant features: Select only the most relevant features to reduce the complexity of the model. Feature engineering: Create new features that are more informative and less prone to overfitting.

Question 36

Where does the data come from for ChatGPT and what form does its training data take?

Answer 36

ChatGPT is trained on a wide variety of data sources, such as the internet, books, websites, articles. The data is typically in text sequences or tokens, and the model learns to predict the next word based on the preceding words.

Question 37

Does ChatGPT use supervised or unsupervised methods to learn and how does this learning happen?

Answer 37

ChatGPT uses supervised learning. Here are the supervised steps to its learning: Data Preparation: A massive dataset of text is collected and preprocessed to remove noise, inconsistencies, and other issues. Tokenisation: The text is broken down into smaller units, such as words or phrases. Model Architecture: A neural network architecture, such as a transformer, is used. Training: During training, the model is fed a sequence of tokens and is asked to predict the next token. The model's predictions are compared to the actual next token, and the model's parameters are adjusted to minimise the error. Fine-tuning: Once the model has been trained on a large dataset, it can be further fine-tuned on specific tasks or domains.

Question 38

Why and how does ChatGPT produce different output texts each time to the same input prompt several times?

Answer 38

predictions, make a comment about form - unsupervised - it learns from training data - llm and picks best prediction from that data. It is taught using training data that involves various examples, and predicts the best fit based on context.

Question 39

What is involved in fine tuning a deep network like GPT-4? How is fine-tuning different from training from scratch?

Answer 39

refer to tranformer models and how GPT4 is a transformer and can be trained on large amounts of information and then fine tuned to focus on one thing

Question 40

What kind of training data is used to fine-tune GPT-4 to encourage ‘helpful’ responses?

Answer 40

Harmful content classifiers

Question 41

Describe three reasons why XAI is important.

Answer 41

Trust, transparency, adjustability (through interpretability and explainability).

Question 42

What components are in a simple convolution neural network?

Answer 42

Input, convolution, pooling, output (fully connected)

Question 43

Describe how a caption can be used to generate an image using a text-to-image AI generator. You can use an analogy.

Answer 43

A caption is the input, therefore the more detail given the better fit the generator will create an image to your idea. A caption to the generator is like a brush to a canvas.

Question 44

Describe one usage of AI text-to-image generation hat could be helpful to society and one that could be harmful.

Answer 44

interior design ideas and sexual exploitation via deepfakes

Question 45

Describe ways to determine whether an image is a deepfake

Answer 45

unnaturally smooth texture, facial deformities, tags, unrealistic depth perception, background blurs to one, details such as too many fingers or clothes with false labels

Question 46

When creating an AI agent for playing the Super Mario game using reinforcement learning, briefly explain how to define the state, action and reward.

Answer 46

State = Current screen Action = movements Reward = win/loss (outcome)

Question 47

State one classification task example and one regression task example

Answer 47

Classification: email spam detection Regression: predicting house prices

Question 48

Predictive AI models, trained using supervised learning, have many applications in healthcare. Describe two such applications, indicating in each case how the predictive model would be trained.

Answer 48

Sensor systems: send signals (e.g, if vitals drop, first responders would be notified), provide summary data, can sensor reminders Systems that predict patient flow in/out, bed use, patient recovery time, organise administrative stuff

Question 49

Describe three ways in which AI technology is currently impacting society, indicating for each whether it is positive or negative.

Answer 49

AI is positively benefiting workers in the workplace as a complementary tool for efficiency. Likewise, AI is a complementary tool for students, positively benefiting their research and idea generation. Although, negatively, AI is breaching ethical concerns with its ability to generate convincing faux videos of people saying or doing things they didn't.

Question 50

describe a regulatory mechanism that is currently being used by governments, or that could be used in the future, to provide some measure of control.

Answer 50

There are regulations around which AI systems hit the market, such as the EU AI laws act which prevents harmful AI models from being disseminated. Further, some models use harmful content classifiers that block users from requesting inappropriate misuse of output.

Question 51

Choose one of the Maori data sovereignty principles that we introduced in lectures. Describe how this principle applies to AI. Discus show you could ensure you follow this principle when building an AI system. Use examples.

Answer 51

Whakapapa: Relationships The principle of whakapapa recognises Maori right to control the collection, storage and distribution of data about Maori - ensuring it positively benefits them - sustainably as well - long term. If data collected about Maori includes under/misrepresentation about Maori individuals/groups then this can perpetuate bias in the model - worsening Maori outcomes from AI. - by ensuring Maori are in control and have full transparency throughout the process of the collection and use of their data. Also, by making sure that the data is representative of Maori, so that in the long run it benefits them rather than indicting bias/stereotypes.

Question 52

What is supervised learning?

Answer 52

The algorithm learns from LABELLED training data and then makes predictions based on the learned relationships.

Question 53

What is unsupervised learning?

Answer 53

The data is UNLABELLED, therefore, the algorithm must find patterns by itself. E.g, regressors in social media - pushes content at us that we have 'interacted' with.

Question 54

How does ML perform tasks?

Answer 54

Maps inputs (training data) to outputs (model) A human writes the ML program, and the ML program creates the model

Question 55

What type of learning was the animal classification in assignment 3?

Answer 55

Supervised - as it is given training data that is labelled (cat, dog, etc with pictures)

Question 56

what are the two main ways of finding patterns in unsupervised learning?

Answer 56

clustering: identifying different types of input dimensionality reduction: simplifying the way data is represented

Question 57

What is an example of clustering (discussed in class)?

Answer 57

A database of 'customer transactions', what items do different types of customers purchase together

Question 58

What is an example of reinforcement learning (discussed in class)?

Answer 58

AI that plays games - AlphaGo etc

Question 59

What is transfer learning?

Answer 59

A model that is trained for one task and re-trained for a new, similar task.

Question 60

Explain tabular data and the areas

Answer 60

Tabular data refers to the structure of training data as a table. Each row is a training instance. (single piece of data) Each instance has many features. Each column holds one of these features. One feature is the output, and the others are inputs.

Question 61

What does feature engineering include and why do we need to do it?

Answer 61

Our data often doesn't arrive in a neat table - we have to create the table. Therefore, we need to create suitable categories, we need to normalise features (make sure one feature doesn't dominate the other), and fix any missing or incorrect data

Question 62

How do we gemoetrically represent datasets with just one feature?

Answer 62

A one-dimensional space - a line

Question 63

How do we geometrically represent a dataset with two features?

Answer 63

Scatterplot

Question 64

How do we geometrically represent a dataset with three features?

Answer 64

A set of points in 3D, using different colours (of dots) for different output labels. The boundaries between output classes can ne thought of as lines or planes, learning often involves finding the right boundary

Question 65

How do ML algorithms represent training items?

Answer 65

As points in a n-dimensional feature space

Question 66

what does Pythagoras theorem refer to in ML algorithms?

Answer 66

the distance between two points in a feature space

Question 67

What is the validation test?

Answer 67

the set we use to explore over/underfitting so we keep the test set for one use only. we can experiment with different parameters on the validation set to see which ones generalise best

Question 68

k=1 and k=5 in k-nearest neighbours will cause what effects?

Answer 68

k=1 will cause overfitting as the model will pay too much attention. to single items k=5 will capture general trends but may ignore actual patterns which is underfitting

Question 69

How does k-NN work?

Answer 69

k-NN is an 'instance-based' classification algorithm that classifies new instances based on similarity (small distance)

Question 70

What are the parts of a decision tree?

Answer 70

a tree is a structure of nodes (boxes) connected by arcs (lines) each node represents one of the features to consider and each nodes arc represent all the possible values for that featuee

Question 71

What is a decision tree?

Answer 71

a classifier that maps input features to an output. it is a structured process that represents a sequence of decisions . they're very easy for humans to understand.

Question 72

how do u train a decision tree?

Answer 72

u pick an input feature and add a node for that feature in the tree split the training set by values of that feature and create arcs to the node for each value at each arc, make a new decision tree with the remaining features

Question 73

What is a binary classifier, how does it work ?

Answer 73

a binary classifier predicts whether an input item is 'in' a class or not. its output is either positive (yes its in the class) or negative (no its not in the class). we can also use a confusion matrix in. a binary classifier to provide extra information about positive and negative predictions

Question 74

How do deepfakes work/what network do they use?

Answer 74

Generative adversarial networks generator network - makes fake images discriminator - judges if an image is fake or real

Question 75

how do u train the discriminator network?

Answer 75

the discriminator classifies both real and fake data from the generator it is then penalised for misclassifying a real instance as fake/fake as real the discriminator then updates its weights through backpropogation

Question 76

how do u train a generator network?

Answer 76

produce generator output from random noise use discriminator to classify this output as a real or fake image penalise generator if classified as fake updates weights through back propagation

Question 77

what is the outcome/goal of training the generator and discriminator?

Answer 77

for the generator to create images so realistic the discriminator cant tell they're fake

Question 78

whats an example from class about when a deepfake was used?

Answer 78

The Gabon deepfake controversy - video released of bongo in 2018

Question 79

What is the kaka AI implementation example in ecology and how does the pipeline work?

Answer 79

Kaka - identification of whether its a new or known individual pipeline: input visuals through data collection then pre processed then object detection then image segmentation then feature matching then knowledgebase then output of known or new individual

Question 80

What is the other AI implementation in ecology example?

Answer 80

modelling plant stress responses : predicting where a species is likely to move and using the ML to understand why we can do this through ML's that predict variables - climate, elevation etc, then training data (NVS surveys) then prediction of area change

Question 81

Any further ways AI is used in ecology?

Answer 81

prediction - using climate and geographical features as inputs to try and predict the abundance changes (species increase or decrease)

Question 82

What are the two types of AI systems in the medical field? what things can they do?

Answer 82

generate and predictive systems medical imaging to diagnose diseases and guide treatment sensing technology - alerts, summary data predicting patient outcomes, patient in/out flow, bed use

Question 83

What are examples of post-hoc approaches?

Answer 83

proxy interpretable models, visualisation, counterfactual explanation, feature importance

Question 84

what does a post-hoc model focus on doing? and how does it do this?

Answer 84

giving an explanation for the model. this can be ether a global (overall model structure/behaviour) or local (how/why the model makes a decision) explanation can also be model specific or model agnostic an explanation can look like a simpler model, visualisation, data points, text explanation, feature importance,

Question 85

In a simplified linear model, what would be of importance? (in terms of feature importance)

Answer 85

coefficients these show the positive/negative relationship and ignore the different scales of the features

Question 86

What is SHAP?

Answer 86

SHAP is Shapley Additive exPlanations this refers to the use of Shapley values where the contribution of each feature tp the overall result is calculates, it uses every possible feature subset, and SHAP approximates Shapley values to give a measure of feature importance

Question 87

what does counterfactual mean?

Answer 87

its causal - if X has not occurred, then Y would not have occurred a counterfactual explanation of a prediction describes the smallest change of the feature values that changes the prediction output

Question 88

What is the purpose of CNNs?

Answer 88

CNN's are for handling image data - to understand spatial relationships in images and avoid the flattening issue (more similar to human vision)

Question 89

What are the parts of a CNN?and what happens between them?

Answer 89

Input (convolutions ->) Feature maps (pooling ->) feature maps (convolutions ->) feature maps (pooling->) feature maps (fully connected ->) output

Question 90

What do the pooling, convolution and fully connected layers of a CNN do?

Answer 90

convolution - has filters, that allow recognition of edges, textures etc each filer slides across the image to create a feature map pooling - reduces dimensionality (width and height) of the feature map this is important for reducing sensitivity, computational cost and gets us closer to the size of the output layer fully connected - connects all neurons from the previous layer to each node in the fully connected layer allows the network to produce an output (final decision/classification)

Question 91

How do u train a CNN?

Answer 91

human designs the network architecture train weights in network (filters and standard weights) with back propagation needs a lot of labelled data - supervised learning

Question 92

Give an analogy for a caption to a text-to-image generator

Answer 92

draft with ideas etc to an essay

Question 93

explain why chagpt produces different responses for the same text

Answer 93

stochastic selection - it chooses an output based on probabilities. when u give ChatGPT a prompt its like ur pointing to a space of its training data and saying I want u to produce text from here! the more detail u give, the more precise the region ur pointing. its given various examples for inputs so it chooses based on probs