Skip to content

Quiz 3: AI Platform Landscape

Test your understanding of commercial and open-source AI platforms.

Questions

Question 1 (Remember)

Which company developed GPT-4 and ChatGPT?

  1. Google
  2. Anthropic
  3. OpenAI
  4. Meta
Answer

C) OpenAI - OpenAI developed the GPT series of models and ChatGPT. Anthropic developed Claude, Google developed Gemini, and Meta developed Llama.

Question 2 (Remember)

What distinguishes Perplexity AI from other AI assistants?

  1. Largest context window
  2. Search-augmented generation with cited sources
  3. Lowest cost
  4. Open-source availability
Answer

B) Search-augmented generation with cited sources - Perplexity AI combines AI text generation with real-time web search, providing answers with cited sources for verification.

Question 3 (Understand)

What does "multimodal" mean in the context of AI models like GPT-4 and Gemini?

  1. Available in multiple languages
  2. Can be accessed through multiple interfaces
  3. Can process multiple types of data (text, images, audio)
  4. Uses multiple neural networks
Answer

C) Can process multiple types of data (text, images, audio) - Multimodal models can handle different input types within a single system, such as analyzing images while generating text responses.

Question 4 (Understand)

Why might an organization choose Claude over ChatGPT for document analysis?

  1. Claude is always cheaper
  2. Claude has more plugins
  3. Claude has a larger context window for long documents
  4. Claude has image generation
Answer

C) Claude has a larger context window for long documents - Claude 3 offers up to 200K tokens context window, making it well-suited for analyzing long documents without chunking.

Question 5 (Apply)

A startup needs to build an AI product but has limited budget. Which platform approach is most appropriate?

  1. Enterprise ChatGPT license
  2. Open-source models like Llama for self-hosting
  3. Multiple proprietary model subscriptions
  4. Custom model training from scratch
Answer

B) Open-source models like Llama for self-hosting - Open-source models have no licensing costs and can be self-hosted, making them ideal for budget-conscious organizations willing to manage infrastructure.

Question 6 (Apply)

You need to generate marketing images from text descriptions. Which platform would you use?

  1. Claude 3 Opus
  2. Perplexity AI
  3. DALL-E or Midjourney
  4. Mistral AI
Answer

C) DALL-E or Midjourney - DALL-E (OpenAI) and Midjourney are specialized text-to-image generation platforms. Claude and Perplexity focus on text, and Mistral focuses on language models.

Question 7 (Analyze)

Compare proprietary models (GPT-4, Claude) with open-source models (Llama, Mistral) in terms of control and capability:

  1. Proprietary always offers more control
  2. Open-source always matches proprietary in capability
  3. Proprietary offers better capability; open-source offers more control
  4. There are no meaningful differences
Answer

C) Proprietary offers better capability; open-source offers more control - Proprietary models generally lead in capability, while open-source models offer more control over deployment, customization, and data privacy.

Question 8 (Analyze)

What is the strategic significance of the Mixtral architecture from Mistral AI?

  1. It's the largest model available
  2. Mixture-of-experts enables efficient inference at scale
  3. It only works on specialized hardware
  4. It requires no training
Answer

B) Mixture-of-experts enables efficient inference at scale - Mixtral uses a mixture-of-experts architecture that activates only relevant model components for each query, providing strong performance with efficient resource usage.

Question 9 (Evaluate)

A healthcare organization needs AI for patient data analysis. Which platform consideration is most critical?

  1. Cost per token
  2. Creative writing capabilities
  3. Data privacy and compliance capabilities
  4. Image generation quality
Answer

C) Data privacy and compliance capabilities - Healthcare involves sensitive patient data subject to regulations like HIPAA. Privacy, security, and compliance are paramount considerations before capability or cost.

Question 10 (Create)

Design a multi-model strategy for an enterprise that needs: (1) customer support chat, (2) long document analysis, (3) image creation, and (4) cost efficiency.

  1. Use GPT-4 for everything
  2. Use open-source for everything
  3. ChatGPT for support, Claude for documents, DALL-E for images, open-source for high-volume/low-complexity
  4. Avoid AI until one model does everything
Answer

C) ChatGPT for support, Claude for documents, DALL-E for images, open-source for high-volume/low-complexity - A multi-model strategy matches platform strengths to use cases, optimizing for capability and cost across different requirements.

Score Interpretation

  • 9-10 correct: Excellent understanding of AI platforms
  • 7-8 correct: Good grasp, review missed concepts
  • 5-6 correct: Fair understanding, revisit chapter sections
  • Below 5: Re-read Chapter 3 before proceeding

Previous: Quiz 2 | Back to Quizzes | Next: Quiz 4 - Prompt Engineering