AI vs Human Game: Can You Tell Who Wrote It?
Free in your browser. No signup, no install.
The AI vs human game is one question on repeat: did a person or an AI write this? You read a short quote with the author hidden, make your call, and the source is revealed. Play it free at app.quoteorbot.com — alone, or screen-shared with a team, class or club. No signup, no install.
What is the AI vs human game?
An AI vs human game puts one piece of content in front of you and asks you to decide where it came from: a person, or a machine. It is the Turing test turned into a round you can play in twenty seconds.
There are a few flavours. In chat-based human-or-bot games you talk to a stranger for a couple of minutes and then guess whether they were a person or a chatbot. In image games you judge whether a photo or painting was generated. In the quote version — the one on this page — you read a single short line of wisdom or motivation with the author hidden and decide: timeless human insight, or a convincing AI imitation?
Quotes turn out to be the sharpest test of the three. They are short enough to read in one breath, they carry no visual artefacts to give the game away, and good AI models are very good at sounding profound. That is exactly what makes the reveal so satisfying — and the wrong guesses so funny.
How to play (solo)
- Open app.quoteorbot.com in any browser, or install the mobile app for iOS or Android.
- Read the quote. The author is hidden.
- Decide: human or AI? On mobile you swipe; in the browser you tap.
- See the reveal — the real author, or confirmation that a model wrote it.
- Repeat. Your accuracy is tracked, so you can see whether your instincts improve.
A round takes seconds, which is why many people play a few quotes a day the way others play Wordle. Over a few dozen rounds you start to build a personal feel for what gives AI writing away.
How to play with a group
The same game becomes a human-or-bot party game the moment you share your screen. It works on Zoom, Google Meet, Microsoft Teams, a projector or a room TV — only the host needs a device.
- Share your screen with app.quoteorbot.com open. Nobody else joins, installs or signs up.
- Read the quote out loud. Hearing the rhythm changes how people judge it.
- Give the room 20 seconds, then vote — hands up, chat reactions or a quick poll: human or AI?
- Ask one person to defend their guess before you reveal. This is what turns a poll into a conversation.
- Reveal and discuss for a minute, then go again.
Two to four rounds fill a five-minute meeting opener. A 45-minute speaking-club session fits four to six rounds with discussion. If you want a full run-of-show for your setting, see the guides for online meetings, ESL classes, speaking clubs and workshops.
Host tip: don't reveal too fast. The twenty seconds of silence before the answer is where people actually think — and where the best arguments start.
Why it's harder than it looks
Most people walk in confident that they can spot a machine. Then an AI line with perfect Stoic cadence fools half the room, and a real quote from a famous author gets voted “obviously a bot”. The game works because both mistakes happen constantly.
After enough rounds, players usually start noticing patterns like these:
Signs a quote might be AI
- A metaphor that is perfectly balanced but says nothing specific.
- Profundity that isn't earned — it sounds wise, but you can't restate the point.
- Symmetrical “not X, but Y” constructions stacked on top of each other.
- Safe, generic nouns: journey, growth, light, path, potential.
Signs a quote might be human
- An odd, concrete detail nobody would invent to sound wise.
- A slightly awkward rhythm or an old-fashioned word.
- An opinion with an edge — humour, irritation, a real stake.
- A line that only makes full sense inside the life of the person who said it.
None of these rules is reliable on its own — which is the point. The game gives you the reps; the instinct builds itself.
AI vs human game formats compared
| Format | What you judge | Time per round | Works for groups? |
|---|---|---|---|
| Quote game (this one) | One short quote, author hidden | 20–60 seconds | Yes — one shared screen |
| Chat bot game | A live chat partner: person or bot | 2–3 minutes | Hard — everyone needs their own device |
| Image game | A photo or artwork: real or generated | 10–30 seconds | Yes, but less to discuss |
| Two truths and a bot | Three statements, one written by AI | 3–5 minutes | Yes — great for teams |
For a group, the quote format wins because everyone judges the same thing at the same time, and a single sentence gives the room something concrete to argue about.
Variations to try
- Two truths and a bot. Everyone writes two true facts about themselves and asks an AI to write a third in their voice. The team guesses which is fake.
- Reverse Turing. Players write their own fake “famous quote” and mix it with real ones. Can the room spot the human forgery?
- Async round. Post a screenshot of the quote card (author hidden) in Slack or Teams with “human or AI?”, then reveal in the thread an hour later.
- Monthly tournament. Keep regular rounds casual, but run a scored tournament once a month with a leaderboard.
- Debate mode. Assign “human” and “AI” sides before the quote appears and let each side argue — perfect for speaking clubs.
Where teams actually use it
We've heard from a remote legal team that plays one round before contract-review calls — everyone arrives a little more sceptical of polished language. A product team uses it as the warm-up before sprint planning. A founder runs a single round at the start of every all-hands as a thirty-second reset. ESL tutors use it because defending a guess forces real grammar: “it must have been…”, “it couldn't have been…”.
The pattern is always the same: almost no setup, a reliable reaction, and a conversation that isn't about the game at all — it's about what makes writing feel human in 2026. If you are looking for more openers like this, see our list of icebreaker games or the this or that questions collection.