Artificial Intelligence Quiz

1. ChatGPT, released by OpenAI in November 2022, reached 100 million users faster than any product in history. How long did it take?

1. One month
2. Six months
3. Two months
4. Two weeks

2. What is 'machine learning' at its core?

1. A branch of mathematics concerned with statistical analysis
2. Programming computers with explicit rules for every task
3. Teaching robots to physically move like humans
4. Training systems to learn patterns from data and improve their performance without being explicitly programmed for each task

3. What is a 'neural network' in AI?

1. A computational model loosely inspired by the human brain, consisting of interconnected layers of nodes that process information
2. A network of scientists collaborating on AI research
3. A type of database optimised for AI applications
4. The internet infrastructure connecting AI servers

4. Which company created the GPT series of language models (GPT-3, GPT-4, etc.)?

1. Google (DeepMind)
2. Meta AI
3. Microsoft
4. OpenAI

5. The 'Turing Test' was proposed by Alan Turing in 1950. What does a machine need to do to pass it?

1. Beat a human chess grandmaster
2. Convince a human evaluator in text conversation that they are talking to a person, not a machine
3. Demonstrate consciousness
4. Solve a mathematical proof faster than any human

6. DeepMind's AlphaFold 2 (2020) was considered a major scientific breakthrough. What did it solve?

1. It created a self-replicating AI code for the first time
2. It created the world's first quantum computer
3. It developed an AI system that passed the Bar Exam
4. It predicted the 3D structure of proteins from amino acid sequences — a 50-year-old grand challenge in biology

7. What is 'hallucination' in the context of AI language models?

1. When an AI confidently generates plausible-sounding but factually incorrect or made-up information
2. When an AI generates different outputs each time it's run
3. When an AI generates dream-like creative content
4. When an AI model becomes confused and stops responding

8. What is 'generative AI' and how does it differ from previous AI applications?

1. AI designed to generate business profits automatically
2. AI that creates new content (text, images, audio, video, code) rather than just classifying or analysing existing data
3. AI that generates electricity from renewable sources
4. AI that generates suggestions for human decision-making only

9. Which company developed the AlphaGo program that defeated world champion Go player Lee Sedol in 2016?

1. DeepMind (Google)
2. IBM
3. Microsoft Research
4. OpenAI

10. What is 'prompt engineering' in the context of AI?

1. Developing the training datasets for AI models
2. Engineering physical devices to work with AI
3. Programming AI models at the hardware level
4. The skill of crafting effective input instructions (prompts) to get the best outputs from AI language models

11. IBM's Watson famously defeated human champions at which TV game show in 2011?

1. Jeopardy!
2. The Price is Right
3. Wheel of Fortune
4. Who Wants to Be a Millionaire?

12. What is 'AGI' — a term frequently discussed in AI circles?

1. Advanced Generative Intelligence — AI that creates content
2. Applied Government Intelligence — AI used in government systems
3. Artificial General Intelligence — a hypothetical AI with human-level (or greater) cognitive abilities across all domains
4. Automated Global Infrastructure for AI computing

13. What is 'reinforcement learning' in AI?

1. A learning paradigm where an AI agent learns by taking actions in an environment and receiving rewards or penalties — optimising for maximum reward
2. Adding more data to reinforce a model's training
3. Teaching an AI by repeatedly correcting its mistakes
4. Using extra computing power to strengthen an AI's neural network

14. Stable Diffusion, Midjourney, and DALL-E are all examples of what type of AI?

1. Image generation models (text-to-image AI)
2. Text generation models
3. Video processing AI
4. Voice synthesis models

15. What does 'LLM' stand for in AI technology?

1. Large Language Model
2. Layered Logic Module
3. Linear Learning Matrix
4. Logic Learning Machine

16. Anthropic, the company that created Claude AI, was founded by former members of which other AI company?

1. DeepMind
2. Google Brain
3. Microsoft Research
4. OpenAI

17. What is a 'transformer' architecture in AI, and why is it significant?

1. A device that transforms electrical voltage for AI servers
2. A neural network architecture introduced in 2017 that underpins most modern LLMs and has revolutionised AI performance on language tasks
3. A type of AI that transforms between different output formats
4. The process of converting data from one format to another

18. What is 'AI alignment' — a major research focus at companies like Anthropic and OpenAI?

1. Aligning AI servers to reduce electromagnetic interference
2. Aligning different AI products from different companies to work together
3. Making AI responses consistent and predictable
4. The challenge of ensuring AI systems behave in ways that are safe, beneficial, and consistent with human values and intentions

19. Google's Gemini (formerly Bard) is a large language model. Which other major Google AI system famously beat human experts at a medical exam?

1. AlphaCode
2. Gemini Ultra
3. Google Health AI
4. MedPaLM 2

20. The 'AI Arms Race' of 2023–2024 primarily involved competition between which major players?

1. IBM and Intel competing for AI chip manufacturing
2. OpenAI/Microsoft, Google/DeepMind, Anthropic, and Meta — competing to release the most capable AI models
3. Universities and corporations competing for AI talent
4. USA and China in a national AI race