Q: What is a modality in AI? Give three examples of different modalities.

A: its basically the type of data an ai uses. like the format. three examples would be text, images and audio.

Q: What is multimodal learning?

A: its when you train the model on multiple modalities at once. so instead of just text it learns from text and pictures together.

Q: Where is multimodal learning used? Can you find one real application and identify the types of information it combines?

A: its used in self driving cars for sure. the car combines images from cameras with 3d spatial data from lidar sensors so it doesn't crash.

Q: What is embodied AI?

A: thats when the ai gets a physical body like a robot. it interacts with the real world instead of just living on a screen.

Q: How are multimodal learning and embodied AI connected?

A: if you want a robot to walk around (embodied), it has to process sights and sounds at the same time to understand its environment, which means it relies heavily on multimodal learning.

Q: What happens when the problem you want to solve becomes too big for any one computer?

A: you just split the problem into pieces and have a bunch of networked computers work on it together.

Q: Suppose 1,000 computers work together. Do you now have one computer that is 1,000 times more powerful? Why or why not?

A: nah not exactly. u lose a lot of time and power just getting the computers to talk to each other and sync up. plus some tasks just cant be split.

Q: Can 1,000 computers agree on something if some of them fail or even lie?

A: yeah they can. its called byzantine fault tolerance. they use voting algorithms so as long as the majority is honest the system keeps working fine.

Q: When you use ChatGPT, Google, Instagram, or an online game, where is the computation actually happening?

A: its happening in huge server farms in data centers far away. basically the cloud.

Q: If you could make millions of computers behave like one dependable machine, what could humanity build that we cannot build today?

A: we could probly simulate crazy things like the entire human brain or predict the weather perfectly for years. stuff that just needs insane computing power.

Q: How did programming evolve from punch cards to high-level languages?

A: we started with literal paper cards with holes, then moved to assembly, and eventually ppl made compilers so we could just write code in plain english-ish text.

Q: What is the justification for having a diverse ecosystem of programming languages?

A: different languages are just better for different things. c++ is for raw speed but python is for writing stuff fast and easy.

Q: What are the key principles of good language design?

A: it needs to be easy to read, safe from dumb memory leaks, and have good libraries so u dont reinvent the wheel.

Q: What are some specific critiques of current tools like Python?

A: python is super slow compared to compiled languages and since it doesnt enforce types you can get some really annoying bugs at runtime.

Q: How do compilers bridge the gap between human-readable code and hardware?

A: they just take the code we type and translate it down into binary 1s and 0s that the cpu actually gets.

Q: What is computational complexity?

A: its a way to measure how much time or memory an algorithm takes as the input gets bigger and bigger.

Q: What exactly is a decision problem?

A: its just a math or logic question that has a strict yes or no answer. like 'is this graph connected?'

Q: What does it mean for a problem to be decidable?

A: it means theres an algorithm out there that can definitely give u the correct yes/no answer in a finite amount of time.

Q: What is the P versus NP problem?

A: P is problems a computer solves fast. NP is problems where checking the answer is fast. the big question is if all NP problems are actually P problems.

Q: Why is there a million-dollar prize associated with proving P vs NP?

A: cause if someone proves P=NP it changes everything. it would break all modern cryptography and let us solve impossible optimization problems.

Q: How do you define AI?

A: artificial intelligence refers to computer systems and machines performing tasks that usually need human intelligence, like logical reasoning, learning from experience, and problem solving.

Q: Can you name at least three different sub-fields of AI?

A: three sub-fields are machine learning for pattern recognition, computer vision for processing images, and natural language processing for understanding human text.

Q: AI has been around for about 70 years so far. Why is it booming right now?

A: its booming now because we finally have massive amounts of data, crazy fast hardware like gpus, and huge algorithmic breakthroughs like deep learning.

Q: Can you name at least three application sectors where robots are being widely employed? What are the reasons?

A: manufacturing uses them for high precision and safety, healthcare uses them for precise surgery and logistics, and agriculture uses them for harvesting and weeding to fight labor shortages.

Q: Can you identify three major challenges for a wheeled autonomous robot performing a 24h surveillance task in a large facility? (e.g., something like Mall of Qatar)

A: power management for 24/7 endurance, navigating dynamic crowds safely, and handling localization drift in huge repeating indoor environments.