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

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 The physical memory chips inside the servers that run the AI model 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

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

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

What is 'hallucination' in the context of AI?

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 When AI produces visual hallucinations or images instead of text Hallucination is a technical term for when AI models become overloaded and crash

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

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

What is the 'basic prompt improvement loop'?

Submit the same prompt three times and choose the best of the three responses Always delete and restart a new conversation if the first response isn't perfect Copy your prompt from online prompt libraries rather than writing your own 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

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

Temperature controls language formality: low for formal writing, high for casual writing 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 Higher temperature always produces better quality — you should maximize it for all tasks Temperature only affects response speed, not content — lower is faster, higher is slower

What usage pattern consistently predicts high-value AI adoption?

Waiting for AI tools to mature further before integrating them into real workflows Trying every new AI tool as soon as it launches to stay current Starting with a specific, recurring real problem; using AI for first drafts with human review; building task-specific prompting habits; knowing which tasks to keep fully human Using AI to replace human involvement in all written tasks entirely

Which of the following is an example of AI that predates ChatGPT and has been running invisibly for years?

A credit card fraud detection system that flags unusual transaction patterns — a classic machine learning application deployed by banks for over a decade ChatGPT Plus, which was the first commercial AI product available to consumers A dropdown menu that shows you your most recently visited websites A basic spreadsheet formula that calculates the sum of a column

How quickly did ChatGPT reach 1 million users after its release in November 2022?

6 months — it grew slowly at first before viral spread ChatGPT has never reached 1 million users — it is primarily used by researchers 1 year — mainstream adoption in consumer AI has historically been slow 5 days — the fastest product in history to that point to cross 1 million users

Even the most capable AI models in 2026 cannot do which of the following?

Summarize a long document into key bullet points Produce grammatically correct text in English Write code that executes correctly in common programming languages Reliably verify whether their own output is factually correct — models can produce false information with the same confidence as true information

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 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 AI tools cost 80% more than traditional software and deliver only 20% of the value

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

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 Because we've had AI winters before, today's AI is definitely not as advanced as claimed AI winters prove that AI never delivers; the current excitement will also disappear

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 There is no risk — AI was trained on legal databases and citations are reliably accurate The risk is low if the AI is asked to cite recent cases from 2024 or later High hallucination risk — AI may generate convincing-looking but nonexistent case citations. Every citation must be verified in a legal database before use

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

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

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

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 Using AI for every task, no matter how simple, to build maximum fluency Delegating final decisions to AI to remove human bias from the process

Which of the following describes a realistic and high-value use of AI in 2026?

Delegating an entire client project to an AI agent with no human review of outputs Using AI to make final legal or medical decisions without a professional reviewing them Using AI to produce a first draft of a client proposal, then editing it with domain knowledge and relationship context the AI doesn't have Using AI to replace all written communication with a client without any human involvement

What does the Gartner Hype Cycle predict about technologies that go through a 'trough of disillusionment'?

Technologies that enter the trough of disillusionment never recover and should be abandoned The trough of disillusionment only applies to hardware technologies, not software like AI Most eventually reach a 'plateau of productivity' where real, consistent value is delivered — the trough represents reality calibration, not permanent failure The Hype Cycle predicts AI will be replaced by a completely different technology by 2027

Why does an AI model perform better on common tasks than on rare or highly specialized knowledge?

AI models are intentionally limited on specialized topics to avoid replacing experts Common tasks appear more frequently in training data, so the model has seen more examples to learn from — rare or specialized topics have fewer training examples, increasing error and hallucination risk Rare topics require more parameters, and models have a fixed number available Common tasks are hard-coded into AI models; rare tasks require internet access

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 Wait until AI tools become more mature and reliable before starting to use them Take a certification course before using any AI tool in a professional context Try every major AI tool simultaneously to quickly determine which is best

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 Telling you whether a news article published yesterday is accurate — AI can't verify real-world current facts 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