Page 355 - Ai_417_V3.0_C9_Flipbook
P. 355
B. Fill in the blanks.
1. An ………………………. learning approach works on an unlabelled dataset.
2. Generative AI could be used to create harmful technologies or weapons, posing significant ………………………. risks.
3. The goal of ………………………. is to find patterns, structures, or representations in data without human interventions.
4. The ethical consideration of ………………………. in generative AI involves the risk of generating biased or discriminatory
content.
5. ………………………. can help in drug discovery by learning and generating molecular structures that have desirable
properties.
6. A major risk associated with generative AI is the potential to create misinformation, such as ………………………. and
deepfakes.
7. Recurrent Neural Networks (RNNs) are especially good at handling ………………………. data like text and music.
8. Generative AI refers to algorithms that create new data resembling ………………………. generated content.
9. The AI tool Artbreeder is used for blending and modifying ………………………. using GANs.
10. ………………………. provides a user-friendly interface for building and training various types of generative models, including
GANs, VAEs, and image classifiers.
C. State whether these statements are true or false.
1. Generative AI can explore new design spaces and optimise systems. ……….……
2. Generative AI is a powerhouse when it comes to creating lots of content in a short time. ……….……
3. The generator in GANs evaluates the generated data to ensure it is realistic. ……….……
4. An important application of RNNs is predictive text input. ……….……
5. Both AE and VAE consist of an encoder and a decoder network. ……….……
6. Runway ML is a platform for creating and deploying Art work and is part of chatGPT. ……….……
7. Generative AI is not capable of composing new music or remixing existing pieces. ……….……
8. Variational Autoencoders (VAEs) are used to sample new data from learned data distributions. ……….……
9. Generative AI algorithms are primarily used to analyse data rather than create new data. ……….……
10. We have no setbacks using Generative AI. ……….……
D. Match the following:
1. Supervised Learning model a. Arts
2. GAN b. Initial input given to the AI model
3. Generative AI c. Generator Network & Discriminator Network
4. Prompt d. Creates noise-free images
5. VAEs e. Labelled
SECTION B (Subjective Type Questions)
A. Short answer type questions:
1. Write one difference between Autoencoders (AEs) and Variational Autoencoders (VAEs).
2. Give two examples of VAEs.
Introduction to Generative AI 353

