AI Basics Quiz: What Do You Actually Know About AI?

What is RLHF (Reinforcement Learning from Human Feedback) and why did it matter for ChatGPT?

RLHF is the process of connecting AI to the live internet for real-time information A fine-tuning technique where human raters scored model responses, training the model to produce outputs humans preferred — turning a text completer into a helpful assistant RLHF is a hardware upgrade that made ChatGPT run faster than previous models RLHF prevents AI from using any data it scraped without permission

What does the history of AI 'winters' and breakthrough cycles tell us about the current AI era?

Because we've had AI winters before, today's AI is definitely not as advanced as claimed AI has had repeated cycles of hype and disappointment — the current era may be different in scale, but calibrated skepticism and attention to real results (not promises) remains the right posture The history of AI winters is irrelevant — current AI is categorically different from all prior AI AI winters prove that AI never delivers; the current excitement will also disappear

What fundamentally changed about AI in late 2022 with ChatGPT's release, compared to AI that existed before?

The 2022 breakthrough was AI becoming conscious and able to truly understand language ChatGPT was the first AI to run on computers; previous AI required specialized hardware AI was invented in 2022 — there was no meaningful AI before ChatGPT Generative AI crossed a quality and accessibility threshold — anyone could use it for genuinely useful work, without technical expertise, for the first time

For which type of task would AI hallucination risk be LOWEST?

Describing a competitor's product features from their latest press release Brainstorming a list of marketing tagline ideas — there are no 'correct' facts to get wrong; creativity and fluency are the objective Looking up the exact revenue of a private company last quarter Finding the specific statute number of a regulation passed in 2025

What happens when a conversation exceeds an AI's context window limit?

The AI loses access to text that falls outside the window — it effectively 'forgets' earlier parts of the conversation and may give inconsistent responses Exceeding the context window triggers an automatic upgrade to a larger model The AI stores the overflow in external memory and retrieves it when needed The AI stops responding and requires a new conversation to be started

What is 'hallucination' in the context of AI?

When AI produces visual hallucinations or images instead of text When AI generates plausible-sounding but factually incorrect information — because the model predicts likely text, not verified facts When AI becomes confused and stops responding to questions Hallucination is a technical term for when AI models become overloaded and crash

Which of the following tasks is AI MOST reliable at?

Calculating the exact compound interest on a complex loan schedule — mathematical precision is a known AI weakness Accurately citing a specific legal case from last month — AI has knowledge cutoffs and hallucination risk with specific facts Producing a well-structured first draft of a professional email — AI draws on vast writing patterns and excels at fluent, structured text generation Telling you whether a news article published yesterday is accurate — AI can't verify real-world current facts

What is the '80/20 trap' in AI adoption?

Only 20% of AI tools are reliable; 80% are experimental and should not be used at work AI tools cost 80% more than traditional software and deliver only 20% of the value AI can only be used for 20% of any professional's tasks; 80% must remain fully human AI handles 80% of a task well but fails at the remaining 20% in ways requiring human correction — making autonomous completion risky, but AI-assisted draft work highly valuable

What is the recommended approach when starting with AI tools for the first time?

Pick one tool and use it consistently on real work for 30 days, building habits through hands-on experimentation rather than reading about AI or switching between tools Take a certification course before using any AI tool in a professional context Wait until AI tools become more mature and reliable before starting to use them Try every major AI tool simultaneously to quickly determine which is best

What made ChatGPT feel fundamentally different from previous AI products like Siri or Alexa?

ChatGPT was a general-purpose system — one interface could write, analyze, code, and answer questions across any domain, unlike assistants built for narrow command-response tasks ChatGPT was the first AI to use the internet in real time to answer questions ChatGPT was different because it ran on quantum computers unavailable to earlier AI ChatGPT was the first AI ever to understand English language

What is the 'basic prompt improvement loop'?

Always delete and restart a new conversation if the first response isn't perfect Submit the same prompt three times and choose the best of the three responses Send a prompt, notice what's missing or wrong in the response, add that context and re-prompt, repeat — treating AI interaction as iterative rather than single-shot Copy your prompt from online prompt libraries rather than writing your own

What is the key difference between 'narrow AI' and 'general-purpose AI'?

Narrow AI runs on phones; general-purpose AI requires large servers Narrow AI is cheaper; general-purpose AI is always more expensive per use Narrow AI is for small companies; general-purpose AI is only for enterprise use Narrow AI is designed for one specific task (spam filtering, recommendations); general-purpose AI handles many different tasks through a single model that users direct with language

What is AI 'hallucination'?

When AI translates text incorrectly because it doesn't know the language When AI becomes confused and stops generating responses When AI generates plausible-sounding but factually incorrect content — stated confidently, without any signal to the user that the information may be wrong When AI produces visual artifacts or distorted images

Which of these everyday products uses machine learning or AI?

A physical thermostat with a timer — it executes pre-set rules without learning A basic alarm clock — it does only what it's explicitly programmed to do Netflix's recommendation engine — it uses machine learning to predict what content you'll enjoy based on your viewing patterns A standard calculator — it follows fixed mathematical rules with no learning

Which of the three habits of effective AI users is most important for working with AI in a professional context?

Delegating final decisions to AI to remove human bias from the process Using AI for every task, no matter how simple, to build maximum fluency Keeping the judgment — AI generates options and drafts; the human remains accountable for decisions and final outputs Never verifying AI output to avoid slowing down the productivity gains

According to McKinsey's 2025 survey, approximately what fraction of organizations report regular AI use in at least one business function?

About 50% — half of organizations have started; half remain uninvolved About 25% — AI adoption is still in early phases for most organizations Less than 10% — AI is still primarily used only by technology companies 88% — AI is now mainstream in organizations, not experimental

A lawyer asks an AI for case citations supporting a specific legal argument. What is the risk, and what should they do?

The risk is only if the lawyer is in a jurisdiction not covered by US law training data High hallucination risk — AI may generate convincing-looking but nonexistent case citations. Every citation must be verified in a legal database before use The risk is low if the AI is asked to cite recent cases from 2024 or later There is no risk — AI was trained on legal databases and citations are reliably accurate

When would you want a LOWER temperature setting for an AI task vs. a HIGHER temperature?

Temperature only affects response speed, not content — lower is faster, higher is slower Higher temperature always produces better quality — you should maximize it for all tasks Lower for precise tasks (data extraction, code, factual answers) where consistency matters; higher for creative tasks (brainstorming, varied options, creative writing) where variation is desirable Temperature controls language formality: low for formal writing, high for casual writing

What does a 'context window' determine in an AI conversation?

The size of the screen where AI output is displayed The time window during which a conversation is saved before being deleted How much text the AI can 'see' and refer to at once — including your messages, shared documents, and the AI's own replies. Text outside the window is effectively forgotten The number of different topics an AI can discuss in a single session

What are 'parameters' in a large language model?

The settings menu where users configure how the AI responds The rules that prevent AI from producing harmful content Billions of numerical values adjusted during training that encode the model's learned patterns — they represent what the model 'knows' about language, facts, and reasoning The physical memory chips inside the servers that run the AI model