Introduction
AI assistants have become part of everyday life, and each major model has developed a recognizable public "personality" shaped by its design and the way people use it. ChatGPT is known as the versatile all-rounder, Gemini for research and Google integration, Claude for careful writing and coding, Llama for open-weight freedom, Midjourney for striking visuals, and DeepSeek for lean, logic-heavy problem solving.
We naturally gravitate toward tools that match how we think. This quiz turns those reputations into six playful archetypes and maps your problem-solving style, work habits, and creative instincts to the AI that fits you best.
Take the 10-question personality quiz first, then explore the personas, see how modern AI actually works, and test yourself with 15 AI trivia questions.
Take the Quiz: Which AI Are You?
Pick the answer that feels most like you, not the one you think sounds best. There are no wrong answers, and your result card shows how strongly you match every persona.
Your result card shows a percentage match for all six personas. A strong second match often explains how you switch styles between work and play.
The 6 AI Personas Explained
Each persona reflects a model’s public reputation, a fun lens rather than a technical benchmark. Most people are a blend, so check your secondary match on your result card too.
| AI Persona | Archetype | Superpower | Ideal Tasks |
|---|---|---|---|
| 🤖 ChatGPT | The All-Rounder Genius | Versatility | Brainstorming, quick explanations, drafting |
| 🔮 Google Gemini | The Multimodal Data Guru | Research synthesis | Web-connected research, mixing text, images and video |
| ✍️ Claude | The Thoughtful Code-Artist | Nuanced craft | Long-form writing, careful code review, analysis |
| 🦙 Meta Llama | The Open-Source Rebel | Independence | Self-hosted projects, custom fine-tuning |
| 🎨 Midjourney | The Creative Visionary | Visual imagination | Concept art, mood boards, branding ideas |
| ⚡ DeepSeek | The Tactical Systems Engineer | Lean efficiency | Math, logic puzzles, algorithm work |
Joseph Weizenbaum was surprised when users, including his own secretary, confided in his simple ELIZA program as if it truly understood them.
1. Meet the AI Models Behind the Personas
The six personas are inspired by real products from very different organizations: OpenAI (ChatGPT), Google DeepMind (Gemini), Anthropic (Claude), Meta (Llama), the independent lab Midjourney, and China’s DeepSeek. How well do you know them?
Round 1: 5 Questions on the Companies & Models Behind Today’s AI
Which company makes ChatGPT?
Show all 5 questions and answers
Which company makes ChatGPT?
Answer: OpenAI
OpenAI released ChatGPT on November 30, 2022. It became one of the fastest-growing consumer apps in history.
Which company develops the Claude family of AI models?
Answer: Anthropic
Anthropic, an AI safety company founded in 2021, released the first Claude model in March 2023.
Google’s Bard chatbot was renamed to what in February 2024?
Answer: Gemini
Google rebranded Bard as Gemini, matching the name of the model family Google DeepMind announced in December 2023.
Meta’s Llama models are notable for being released as what?
Answer: Open-weight models that developers can download and run
Meta has released Llama model weights for developers to download, fine-tune, and run on their own hardware under its license, which fuels a large open-model ecosystem.
Midjourney, the popular AI image generator, originally ran mainly through which platform?
Answer: Discord
Midjourney launched its open beta in 2022 as a Discord bot. Users typed prompts in chat channels, before the company later added a web interface.
The 2017 Transformer paper from Google researchers underpins GPT, Gemini, Claude, Llama, and DeepSeek alike.
2. How Modern AI Actually Works
Today’s chat assistants are large language models built on the Transformer architecture. They are trained on vast amounts of text to predict what comes next, then refined with techniques such as human feedback to be more helpful and safe.
Understanding the basics, including hallucinations, multimodality, and fine-tuning, helps you use any AI tool more effectively and know when to double-check its answers.
Round 2: 5 Questions on Transformers, Hallucinations & Training
What does "GPT" stand for?
Show all 5 questions and answers
What does "GPT" stand for?
Answer: Generative Pre-trained Transformer
GPT models are generative (they produce text), pre-trained on large amounts of data, and built on the Transformer architecture.
The 2017 paper that introduced the Transformer architecture is titled:
Answer: Attention Is All You Need
Written by researchers at Google, "Attention Is All You Need" introduced the self-attention-based Transformer, the foundation of nearly every modern large language model.
When an AI chatbot confidently states something false or made up, this is commonly called:
Answer: A hallucination
Language models predict plausible text, which can produce fluent but incorrect answers. Checking important facts against reliable sources is always wise.
What does RLHF, a common technique for making chatbots more helpful, stand for?
Answer: Reinforcement Learning from Human Feedback
In RLHF, people rate or compare model responses, and the model is trained to prefer the kinds of answers humans judge as better.
An AI model that can work with several input types, such as text, images and audio, is described as:
Answer: Multimodal
Multimodal models can read an image, listen to audio, or analyze a chart alongside text. Most major assistants now offer some multimodal features.
In 2024, the Physics Nobel went to John Hopfield and Geoffrey Hinton for neural network foundations, and part of the Chemistry Nobel went to Demis Hassabis and John Jumper for AlphaFold.
AI Milestones Timeline
Today’s chatbots stand on more than 70 years of research. These moments shaped the field:
| Milestone | Year | Why It Mattered |
|---|---|---|
| Turing proposes the "imitation game" | 1950 | Framed the question "Can machines think?" |
| "Artificial intelligence" coined for Dartmouth workshop | 1955–56 | Launched AI as a research field |
| ELIZA chatbot at MIT | 1966 | Showed how readily people attribute understanding to machines |
| Deep Blue defeats Garry Kasparov | 1997 | First win by a computer over a reigning world chess champion in a match |
| AlphaGo defeats Lee Sedol | 2016 | Mastered Go, long considered too complex for computers |
| "Attention Is All You Need" introduces the Transformer | 2017 | The architecture behind modern large language models |
| ChatGPT launches | 2022 | Brought conversational AI to a mass audience |
3. AI History & Breakthrough Moments
From Turing’s thought experiment to machines beating world champions at chess and Go, AI history is full of surprises. Test how well you know the turning points.
Round 3: 5 Questions on Turing, Deep Blue, AlphaGo & More
Who proposed the "imitation game," now known as the Turing test, in 1950?
Show all 5 questions and answers
Who proposed the "imitation game," now known as the Turing test, in 1950?
Answer: Alan Turing
Alan Turing’s paper "Computing Machinery and Intelligence" asked whether a machine could convincingly imitate a human in conversation.
Who coined the term "artificial intelligence" in the proposal for the 1956 Dartmouth workshop?
Answer: John McCarthy
John McCarthy, together with Minsky, Nathaniel Rochester, and Claude Shannon, wrote the 1955 proposal for the Dartmouth summer project that launched AI as a field.
In 1997, IBM’s Deep Blue defeated which world chess champion in a match?
Answer: Garry Kasparov
Deep Blue beat Garry Kasparov 3.5–2.5 in a six-game rematch, the first match loss by a reigning world champion to a computer under standard time controls.
Google DeepMind’s AlphaGo defeated Go champion Lee Sedol in 2016 by what score?
Answer: 4–1
AlphaGo won four of five games in Seoul. Lee Sedol’s win in game four, sparked by his famous "move 78," was the only game a human took.
In January 2025, the release of which company’s R1 reasoning model rattled AI stocks?
Answer: DeepSeek
DeepSeek, based in Hangzhou, China, released R1 as an open-weight model that performed strongly at a reportedly low training cost. Nvidia’s shares fell about 17% in a single day.
Fun Facts About AI
Surprising facts to share with fellow AI fans:
Analysts estimated ChatGPT reached around 100 million monthly users within about two months of launch, a record for consumer apps at the time.
In game two against Lee Sedol, AlphaGo played a move so unusual that commentators first thought it was a mistake. It is now one of the most famous moves in Go history.
Many image generators use diffusion models, which start from random static and refine it step by step into a picture that matches your prompt.
"Artificial intelligence" was coined in a 1955 funding proposal, a year before the Dartmouth workshop that officially launched the field.