I Asked AIs “What is AI?” and Then Had Them Rate Each Other’s Answers

Have you ever wondered how an artificial intelligence would define itself if you asked, “What is AI?” I was curious, so I conducted a little experiment.

How did the experiment work?

The rules were simple:

I asked them all the same question: “What is AI?”

For each AI, I used its most capable “reasoning” or “thinking” model available for free.

I collected the responses and asked the AIs to rate each answer on a scale of 1 to 10—including their own.

The goal wasn’t necessarily to determine which one was “better,” but rather to see how one AI evaluates the response of another.

Claude received the highest score.

Gemini was the most generous, awarding the highest scores.

Of course, this isn’t a scientific study. It’s based on a single question, a single trial, and the subjective judgment of the AIs themselves. Still, it was quite fascinating to see how these tools perceive one another. You could try asking the same question yourself—you might get different results!

Frequently Asked Questions

What is AI?

Artificial intelligence (AI) is a field of computer science focused on creating systems that can perform tasks commonly associated with human intelligence, such as understanding language, recognizing patterns, generating content, and solving problems.

How can AI systems be compared?

AI systems can be compared by asking them the same questions and evaluating their answers using consistent criteria, such as accuracy, clarity, relevance, completeness, and reasoning quality. A single comparison, however, cannot establish an overall ranking.

Can an AI evaluate its own answer?

Yes. An AI system can be asked to assess its own response against defined criteria or a scoring scale. However, self-evaluation is subjective and should not be treated as independent proof of quality.

Which AI received the highest score in this experiment?

Claude received the highest score in this particular experiment. This result reflects the ratings collected in one trial and does not establish that Claude is always better than other AI systems.

Which AI was the most generous when rating other answers?

Gemini was the most generous evaluator in this experiment, awarding the highest scores among the participating AI systems.

Is this AI comparison a scientific study?

No. It was an informal experiment based on one question, one trial, and subjective AI-generated ratings. More questions, repeated trials, human evaluation, and standardized criteria would be needed for a more reliable comparison.