AI and technology

Is social media more harmful or helpful?

Position tracked against: “Social media is more harmful than helpful.”

All datesLast 30 daysLast 7 days

How do chatbots’ citation patterns change over time?

Share of answers with a citation, each day.

0%25%50%75%100%Sep 11Sep 28Oct 15Nov 1GPT-5, Sep 11: 0% (n = 1)GPT-5, Sep 12: 0% (n = 1)GPT-5, Sep 13: 0% (n = 1)GPT-5, Sep 14: 0% (n = 1)GPT-5, Sep 15: 0% (n = 1)GPT-5, Sep 16: 0% (n = 1)GPT-5, Sep 17: 0% (n = 1)GPT-5, Sep 18: 0% (n = 1)GPT-5, Sep 19: 0% (n = 1)GPT-5, Sep 20: 0% (n = 1)GPT-5, Sep 21: 0% (n = 1)GPT-5, Sep 22: 0% (n = 1)GPT-5, Sep 23: 0% (n = 1)GPT-5, Sep 24: 0% (n = 1)GPT-5, Sep 25: 0% (n = 1)GPT-5, Sep 26: 0% (n = 1)GPT-5, Sep 27: 0% (n = 1)GPT-5, Sep 28: 0% (n = 1)GPT-5, Sep 29: 0% (n = 1)GPT-5, Sep 30: 0% (n = 1)GPT-5, Oct 1: 0% (n = 1)GPT-5, Oct 2: 0% (n = 1)GPT-5, Oct 3: 0% (n = 1)GPT-5, Oct 4: 0% (n = 1)GPT-5, Oct 5: 0% (n = 1)GPT-5, Oct 6: 0% (n = 1)GPT-5, Oct 7: 0% (n = 1)GPT-5, Oct 8: 0% (n = 1)GPT-5, Oct 9: 0% (n = 1)GPT-5, Oct 10: 0% (n = 1)GPT-5, Oct 11: 0% (n = 1)GPT-5, Oct 12: 0% (n = 1)GPT-5, Oct 13: 0% (n = 1)GPT-5, Oct 14: 0% (n = 1)GPT-5, Oct 15: 0% (n = 1)GPT-5, Oct 16: 0% (n = 1)GPT-5, Oct 17: 0% (n = 1)GPT-5, Oct 18: 0% (n = 1)GPT-5, Oct 19: 0% (n = 1)GPT-5, Oct 20: 0% (n = 1)GPT-5, Oct 21: 0% (n = 1)GPT-5, Oct 22: 0% (n = 1)GPT-5, Oct 23: 0% (n = 1)GPT-5, Oct 24: 0% (n = 1)GPT-5, Oct 25: 0% (n = 1)GPT-5, Oct 26: 0% (n = 1)GPT-5, Oct 27: 0% (n = 1)GPT-5, Oct 28: 0% (n = 1)GPT-5, Oct 29: 0% (n = 1)GPT-5, Oct 30: 0% (n = 1)GPT-5, Oct 31: 0% (n = 1)GPT-5, Nov 1: 0% (n = 1)GPT-5, Nov 2: 0% (n = 1)GPT-5, Nov 3: 0% (n = 1)GPT-5, Nov 4: 0% (n = 1)GPT-5, Nov 5: 0% (n = 1)GPT-5, Nov 6: 0% (n = 1)GPT-5, Nov 7: 0% (n = 1)GPT-5, Nov 8: 0% (n = 1)Claude Sonnet 5, Sep 11: 100% (n = 1)Claude Sonnet 5, Sep 12: 0% (n = 1)Claude Sonnet 5, Sep 13: 100% (n = 1)Claude Sonnet 5, Sep 14: 100% (n = 1)Claude Sonnet 5, Sep 15: 100% (n = 1)Claude Sonnet 5, Sep 16: 0% (n = 1)Claude Sonnet 5, Sep 17: 100% (n = 1)Claude Sonnet 5, Sep 18: 100% (n = 1)Claude Sonnet 5, Sep 19: 100% (n = 1)Claude Sonnet 5, Sep 20: 0% (n = 1)Claude Sonnet 5, Sep 21: 100% (n = 1)Claude Sonnet 5, Sep 22: 100% (n = 1)Claude Sonnet 5, Sep 23: 100% (n = 1)Claude Sonnet 5, Sep 24: 100% (n = 1)Claude Sonnet 5, Sep 25: 100% (n = 1)Claude Sonnet 5, Sep 26: 100% (n = 1)Claude Sonnet 5, Sep 27: 100% (n = 1)Claude Sonnet 5, Sep 28: 100% (n = 1)Claude Sonnet 5, Sep 29: 100% (n = 1)Claude Sonnet 5, Sep 30: 100% (n = 1)Claude Sonnet 5, Oct 1: 100% (n = 1)Claude Sonnet 5, Oct 2: 100% (n = 1)Claude Sonnet 5, Oct 3: 100% (n = 1)Claude Sonnet 5, Oct 4: 100% (n = 1)Claude Sonnet 5, Oct 5: 0% (n = 1)Claude Sonnet 5, Oct 6: 0% (n = 1)Claude Sonnet 5, Oct 7: 100% (n = 1)Claude Sonnet 5, Oct 8: 0% (n = 1)Claude Sonnet 5, Oct 9: 100% (n = 1)Claude Sonnet 5, Oct 10: 100% (n = 1)Claude Sonnet 5, Oct 11: 0% (n = 1)Claude Sonnet 5, Oct 12: 100% (n = 1)Claude Sonnet 5, Oct 13: 100% (n = 1)Claude Sonnet 5, Oct 14: 100% (n = 1)Claude Sonnet 5, Oct 15: 100% (n = 1)Claude Sonnet 5, Oct 16: 100% (n = 1)Claude Sonnet 5, Oct 17: 100% (n = 1)Claude Sonnet 5, Oct 18: 0% (n = 1)Claude Sonnet 5, Oct 19: 0% (n = 1)Claude Sonnet 5, Oct 20: 100% (n = 1)Claude Sonnet 5, Oct 21: 100% (n = 1)Claude Sonnet 5, Oct 22: 100% (n = 1)Claude Sonnet 5, Oct 23: 100% (n = 1)Claude Sonnet 5, Oct 24: 0% (n = 1)Claude Sonnet 5, Oct 25: 100% (n = 1)Claude Sonnet 5, Oct 26: 100% (n = 1)Claude Sonnet 5, Oct 27: 0% (n = 1)Claude Sonnet 5, Oct 28: 100% (n = 1)Claude Sonnet 5, Oct 29: 100% (n = 1)Claude Sonnet 5, Oct 30: 100% (n = 1)Claude Sonnet 5, Oct 31: 100% (n = 1)Claude Sonnet 5, Nov 1: 100% (n = 1)Claude Sonnet 5, Nov 2: 100% (n = 1)Claude Sonnet 5, Nov 3: 0% (n = 1)Claude Sonnet 5, Nov 4: 100% (n = 1)Claude Sonnet 5, Nov 5: 0% (n = 1)Claude Sonnet 5, Nov 6: 0% (n = 1)Claude Sonnet 5, Nov 7: 0% (n = 1)Claude Sonnet 5, Nov 8: 0% (n = 1)Gemini 3.5 Flash, Sep 11: 100% (n = 1)Gemini 3.5 Flash, Sep 12: 100% (n = 1)Gemini 3.5 Flash, Sep 13: 100% (n = 1)Gemini 3.5 Flash, Sep 14: 100% (n = 1)Gemini 3.5 Flash, Sep 15: 100% (n = 1)Gemini 3.5 Flash, Sep 16: 100% (n = 1)Gemini 3.5 Flash, Sep 17: 100% (n = 1)Gemini 3.5 Flash, Sep 18: 0% (n = 1)Gemini 3.5 Flash, Sep 19: 0% (n = 1)Gemini 3.5 Flash, Sep 20: 100% (n = 1)Gemini 3.5 Flash, Sep 21: 0% (n = 1)Gemini 3.5 Flash, Sep 22: 100% (n = 1)Gemini 3.5 Flash, Sep 23: 100% (n = 1)Gemini 3.5 Flash, Sep 24: 0% (n = 1)Gemini 3.5 Flash, Sep 25: 100% (n = 1)Gemini 3.5 Flash, Sep 26: 100% (n = 1)Gemini 3.5 Flash, Sep 27: 0% (n = 1)Gemini 3.5 Flash, Sep 28: 100% (n = 1)Gemini 3.5 Flash, Sep 29: 100% (n = 1)Gemini 3.5 Flash, Sep 30: 100% (n = 1)Gemini 3.5 Flash, Oct 1: 0% (n = 1)Gemini 3.5 Flash, Oct 2: 0% (n = 1)Gemini 3.5 Flash, Oct 3: 0% (n = 1)Gemini 3.5 Flash, Oct 4: 100% (n = 1)Gemini 3.5 Flash, Oct 5: 0% (n = 1)Gemini 3.5 Flash, Oct 6: 100% (n = 1)Gemini 3.5 Flash, Oct 7: 0% (n = 1)Gemini 3.5 Flash, Oct 8: 100% (n = 1)Gemini 3.5 Flash, Oct 9: 100% (n = 1)Gemini 3.5 Flash, Oct 10: 100% (n = 1)Gemini 3.5 Flash, Oct 11: 100% (n = 1)Gemini 3.5 Flash, Oct 12: 100% (n = 1)Gemini 3.5 Flash, Oct 13: 100% (n = 1)Gemini 3.5 Flash, Oct 14: 100% (n = 1)Gemini 3.5 Flash, Oct 15: 100% (n = 1)Gemini 3.5 Flash, Oct 16: 100% (n = 1)Gemini 3.5 Flash, Oct 17: 100% (n = 1)Gemini 3.5 Flash, Oct 18: 100% (n = 1)Gemini 3.5 Flash, Oct 19: 100% (n = 1)Gemini 3.5 Flash, Oct 20: 100% (n = 1)Gemini 3.5 Flash, Oct 21: 100% (n = 1)Gemini 3.5 Flash, Oct 22: 100% (n = 1)Gemini 3.5 Flash, Oct 23: 100% (n = 1)Gemini 3.5 Flash, Oct 24: 100% (n = 1)Gemini 3.5 Flash, Oct 25: 100% (n = 1)Gemini 3.5 Flash, Oct 26: 100% (n = 1)Gemini 3.5 Flash, Oct 27: 100% (n = 1)Gemini 3.5 Flash, Oct 28: 100% (n = 1)Gemini 3.5 Flash, Oct 29: 100% (n = 1)Gemini 3.5 Flash, Oct 30: 100% (n = 1)Gemini 3.5 Flash, Oct 31: 100% (n = 1)Gemini 3.5 Flash, Nov 1: 100% (n = 1)Gemini 3.5 Flash, Nov 2: 100% (n = 1)Gemini 3.5 Flash, Nov 3: 100% (n = 1)Gemini 3.5 Flash, Nov 4: 100% (n = 1)Gemini 3.5 Flash, Nov 5: 100% (n = 1)Gemini 3.5 Flash, Nov 6: 100% (n = 1)Gemini 3.5 Flash, Nov 7: 100% (n = 1)Gemini 3.5 Flash, Nov 8: 100% (n = 1)Grok 4.5, Sep 11: 100% (n = 1)Grok 4.5, Sep 12: 0% (n = 1)Grok 4.5, Sep 13: 100% (n = 1)Grok 4.5, Sep 14: 0% (n = 1)Grok 4.5, Sep 15: 0% (n = 1)Grok 4.5, Sep 16: 100% (n = 1)Grok 4.5, Sep 17: 100% (n = 1)Grok 4.5, Sep 18: 100% (n = 1)Grok 4.5, Sep 19: 100% (n = 1)Grok 4.5, Sep 20: 100% (n = 1)Grok 4.5, Sep 21: 100% (n = 1)Grok 4.5, Sep 22: 100% (n = 1)Grok 4.5, Sep 23: 0% (n = 1)Grok 4.5, Sep 24: 100% (n = 1)Grok 4.5, Sep 25: 100% (n = 1)Grok 4.5, Sep 26: 100% (n = 1)Grok 4.5, Sep 27: 100% (n = 1)Grok 4.5, Sep 28: 100% (n = 1)Grok 4.5, Sep 29: 0% (n = 1)Grok 4.5, Sep 30: 100% (n = 1)Grok 4.5, Oct 1: 100% (n = 1)Grok 4.5, Oct 2: 100% (n = 1)Grok 4.5, Oct 3: 100% (n = 1)Grok 4.5, Oct 4: 100% (n = 1)Grok 4.5, Oct 5: 100% (n = 1)Grok 4.5, Oct 6: 100% (n = 1)Grok 4.5, Oct 7: 100% (n = 1)Grok 4.5, Oct 8: 100% (n = 1)Grok 4.5, Oct 9: 100% (n = 1)Grok 4.5, Oct 10: 100% (n = 1)Grok 4.5, Oct 11: 100% (n = 1)Grok 4.5, Oct 12: 100% (n = 1)Grok 4.5, Oct 13: 100% (n = 1)Grok 4.5, Oct 14: 100% (n = 1)Grok 4.5, Oct 15: 100% (n = 1)Grok 4.5, Oct 16: 100% (n = 1)Grok 4.5, Oct 17: 100% (n = 1)Grok 4.5, Oct 18: 100% (n = 1)Grok 4.5, Oct 19: 100% (n = 1)Grok 4.5, Oct 20: 100% (n = 1)Grok 4.5, Oct 21: 100% (n = 1)Grok 4.5, Oct 22: 100% (n = 1)Grok 4.5, Oct 23: 100% (n = 1)Grok 4.5, Oct 24: 100% (n = 1)Grok 4.5, Oct 25: 100% (n = 1)Grok 4.5, Oct 26: 100% (n = 1)Grok 4.5, Oct 27: 100% (n = 1)Grok 4.5, Oct 28: 100% (n = 1)Grok 4.5, Oct 29: 100% (n = 1)Grok 4.5, Oct 30: 100% (n = 1)Grok 4.5, Oct 31: 100% (n = 1)Grok 4.5, Nov 1: 100% (n = 1)Grok 4.5, Nov 2: 100% (n = 1)Grok 4.5, Nov 3: 100% (n = 1)Grok 4.5, Nov 4: 100% (n = 1)Grok 4.5, Nov 5: 100% (n = 1)Grok 4.5, Nov 6: 100% (n = 1)Grok 4.5, Nov 7: 100% (n = 1)Grok 4.5, Nov 8: 100% (n = 1)Muse Spark 1.2, Sep 11: 100% (n = 1)Muse Spark 1.2, Sep 12: 100% (n = 1)Muse Spark 1.2, Sep 13: 100% (n = 1)Muse Spark 1.2, Sep 14: 100% (n = 1)Muse Spark 1.2, Sep 15: 100% (n = 1)Muse Spark 1.2, Sep 16: 100% (n = 1)Muse Spark 1.2, Sep 17: 100% (n = 1)Muse Spark 1.2, Sep 18: 100% (n = 1)Muse Spark 1.2, Sep 19: 100% (n = 1)Muse Spark 1.2, Sep 20: 100% (n = 1)Muse Spark 1.2, Sep 21: 100% (n = 1)Muse Spark 1.2, Sep 22: 100% (n = 1)Muse Spark 1.2, Sep 23: 100% (n = 1)Muse Spark 1.2, Sep 24: 100% (n = 1)Muse Spark 1.2, Sep 25: 100% (n = 1)Muse Spark 1.2, Sep 26: 100% (n = 1)Muse Spark 1.2, Sep 27: 100% (n = 1)Muse Spark 1.2, Sep 28: 100% (n = 1)Muse Spark 1.2, Sep 29: 100% (n = 1)Muse Spark 1.2, Sep 30: 100% (n = 1)Muse Spark 1.2, Oct 1: 100% (n = 1)Muse Spark 1.2, Oct 2: 100% (n = 1)Muse Spark 1.2, Oct 3: 100% (n = 1)Muse Spark 1.2, Oct 4: 100% (n = 1)Muse Spark 1.2, Oct 5: 100% (n = 1)Muse Spark 1.2, Oct 6: 100% (n = 1)Muse Spark 1.2, Oct 7: 100% (n = 1)Muse Spark 1.2, Oct 8: 100% (n = 1)Muse Spark 1.2, Oct 9: 100% (n = 1)Muse Spark 1.2, Oct 10: 100% (n = 1)Muse Spark 1.2, Oct 11: 100% (n = 1)Muse Spark 1.2, Oct 12: 100% (n = 1)Muse Spark 1.2, Oct 13: 100% (n = 1)Muse Spark 1.2, Oct 14: 100% (n = 1)Muse Spark 1.2, Oct 15: 100% (n = 1)Muse Spark 1.2, Oct 16: 0% (n = 1)Muse Spark 1.2, Oct 17: 100% (n = 1)Muse Spark 1.2, Oct 18: 100% (n = 1)Muse Spark 1.2, Oct 19: 0% (n = 1)Muse Spark 1.2, Oct 20: 100% (n = 1)Muse Spark 1.2, Oct 21: 100% (n = 1)Muse Spark 1.2, Oct 22: 100% (n = 1)Muse Spark 1.2, Oct 23: 100% (n = 1)Muse Spark 1.2, Oct 24: 100% (n = 1)Muse Spark 1.2, Oct 25: 100% (n = 1)Muse Spark 1.2, Oct 26: 100% (n = 1)Muse Spark 1.2, Oct 27: 100% (n = 1)Muse Spark 1.2, Oct 28: 100% (n = 1)Muse Spark 1.2, Oct 29: 100% (n = 1)Muse Spark 1.2, Oct 30: 100% (n = 1)Muse Spark 1.2, Oct 31: 100% (n = 1)Muse Spark 1.2, Nov 1: 100% (n = 1)Muse Spark 1.2, Nov 2: 100% (n = 1)Muse Spark 1.2, Nov 3: 100% (n = 1)Muse Spark 1.2, Nov 4: 100% (n = 1)Muse Spark 1.2, Nov 5: 100% (n = 1)Muse Spark 1.2, Nov 6: 100% (n = 1)Muse Spark 1.2, Nov 7: 100% (n = 1)Muse Spark 1.2, Nov 8: 100% (n = 1)
GPT-5Claude Sonnet 5Gemini 3.5 FlashGrok 4.5Muse Spark 1.2
Data

Methodology

How do chatbots’ citation patterns change over time?

Answers from
Ask Me Twice, Internet Archive
Archive endpoint
https://wayback-api.archive.org/services/amt/api/v1/responses/?detail=full
Model endpoint
https://openrouter.ai/api/v1/chat/completions (the archive asks through OpenRouter)
Web search
On from Sep 11, 2026
Collection
Once a day, one answer per chatbot and question, each in a new chat with no history
Dates
Sep 11, 2026 to Nov 8, 2026
Answers
295
Models
  • GPT-5 (openai/gpt-5)
  • Claude Sonnet 5 (anthropic/claude-sonnet-5)
  • Gemini 3.5 Flash (google/gemini-3.5-flash)
  • Grok 4.5 (x-ai/grok-4.5)
  • Muse Spark 1.2 (meta/muse-spark-1.2)
Prompt
Is social media more harmful or helpful?

Position over time

Average position each day on one yes-or-no question (+1 yes, −1 no).

-1.00-0.500.00+0.50+1.00Sep 11Oct 5Oct 27GPT-5, Sep 11: 0.00 (n = 1)GPT-5, Sep 12: 0.00 (n = 1)GPT-5, Sep 13: 0.00 (n = 1)GPT-5, Sep 14: 0.00 (n = 1)GPT-5, Sep 15: 0.00 (n = 1)GPT-5, Sep 16: 0.00 (n = 1)GPT-5, Sep 23: 0.00 (n = 1)GPT-5, Oct 5: 0.00 (n = 1)GPT-5, Oct 16: 0.00 (n = 1)GPT-5, Oct 27: 0.00 (n = 1)GPT-5, Oct 28: 0.00 (n = 1)GPT-5, Oct 29: 0.00 (n = 1)Claude Sonnet 5, Sep 11: 0.00 (n = 1)Claude Sonnet 5, Sep 12: 0.00 (n = 1)Claude Sonnet 5, Sep 13: 0.00 (n = 1)Claude Sonnet 5, Sep 14: 0.00 (n = 1)Claude Sonnet 5, Sep 15: 0.00 (n = 1)Claude Sonnet 5, Sep 16: 0.00 (n = 1)Claude Sonnet 5, Sep 23: 0.00 (n = 1)Claude Sonnet 5, Oct 5: 0.00 (n = 1)Claude Sonnet 5, Oct 16: 0.00 (n = 1)Claude Sonnet 5, Oct 27: 0.00 (n = 1)Claude Sonnet 5, Oct 28: 0.00 (n = 1)Claude Sonnet 5, Oct 29: 0.00 (n = 1)Gemini 3.5 Flash, Sep 11: 0.00 (n = 1)Gemini 3.5 Flash, Sep 12: 0.00 (n = 1)Gemini 3.5 Flash, Sep 13: 0.00 (n = 1)Gemini 3.5 Flash, Sep 14: 0.00 (n = 1)Gemini 3.5 Flash, Sep 15: 0.00 (n = 1)Gemini 3.5 Flash, Sep 16: 0.00 (n = 1)Gemini 3.5 Flash, Sep 23: 0.00 (n = 1)Gemini 3.5 Flash, Oct 5: 0.00 (n = 1)Gemini 3.5 Flash, Oct 16: +0.50 (n = 1)Gemini 3.5 Flash, Oct 27: 0.00 (n = 1)Gemini 3.5 Flash, Oct 28: 0.00 (n = 1)Gemini 3.5 Flash, Oct 29: +0.50 (n = 1)Gemini 3.5 Flash, Oct 30: 0.00 (n = 1)Grok 4.5, Sep 11: 0.00 (n = 1)Grok 4.5, Sep 12: 0.00 (n = 1)Grok 4.5, Sep 13: 0.00 (n = 1)Grok 4.5, Sep 14: 0.00 (n = 1)Grok 4.5, Sep 15: 0.00 (n = 1)Grok 4.5, Sep 16: 0.00 (n = 1)Grok 4.5, Sep 17: -0.50 (n = 1)Grok 4.5, Sep 23: 0.00 (n = 1)Grok 4.5, Oct 5: 0.00 (n = 1)Grok 4.5, Oct 16: 0.00 (n = 1)Grok 4.5, Oct 27: 0.00 (n = 1)Grok 4.5, Oct 28: 0.00 (n = 1)Grok 4.5, Oct 29: 0.00 (n = 1)Muse Spark 1.2, Sep 11: 0.00 (n = 1)Muse Spark 1.2, Sep 12: 0.00 (n = 1)Muse Spark 1.2, Sep 13: 0.00 (n = 1)Muse Spark 1.2, Sep 14: 0.00 (n = 1)Muse Spark 1.2, Sep 15: 0.00 (n = 1)Muse Spark 1.2, Sep 16: 0.00 (n = 1)Muse Spark 1.2, Sep 17: 0.00 (n = 1)Muse Spark 1.2, Sep 21: 0.00 (n = 1)Muse Spark 1.2, Sep 23: 0.00 (n = 1)Muse Spark 1.2, Oct 5: 0.00 (n = 1)Muse Spark 1.2, Oct 16: 0.00 (n = 1)Muse Spark 1.2, Oct 27: 0.00 (n = 1)Muse Spark 1.2, Oct 28: 0.00 (n = 1)Muse Spark 1.2, Oct 29: 0.00 (n = 1)
GPT-5Claude Sonnet 5Gemini 3.5 FlashGrok 4.5Muse Spark 1.2

Social media is more harmful than helpful.

Data

Methodology

Position over time

Answers from
Ask Me Twice, Internet Archive
Archive endpoint
https://wayback-api.archive.org/services/amt/api/v1/responses/?detail=full
Model endpoint
https://openrouter.ai/api/v1/chat/completions (the archive asks through OpenRouter)
Web search
On from Sep 11, 2026
Collection
Once a day, one answer per chatbot and question, each in a new chat with no history
Dates
Sep 11, 2026 to Nov 8, 2026
Answers
295
Models
  • GPT-5 (openai/gpt-5)
  • Claude Sonnet 5 (anthropic/claude-sonnet-5)
  • Gemini 3.5 Flash (google/gemini-3.5-flash)
  • Grok 4.5 (x-ai/grok-4.5)
  • Muse Spark 1.2 (meta/muse-spark-1.2)
Prompt
Is social media more harmful or helpful?
Annotator
claude-sonnet-5-5, prompt v1
Position scale
Yes +1, leans yes +0.5, both sides 0, leans no −0.5, no −1. Declines are left out.

How often do chatbots cite external sources?

Share of answers with at least one link/citation.

Data

Methodology

How often do chatbots cite external sources?

Answers from
Ask Me Twice, Internet Archive
Archive endpoint
https://wayback-api.archive.org/services/amt/api/v1/responses/?detail=full
Model endpoint
https://openrouter.ai/api/v1/chat/completions (the archive asks through OpenRouter)
Web search
On from Sep 11, 2026
Collection
Once a day, one answer per chatbot and question, each in a new chat with no history
Dates
Sep 11, 2026 to Nov 8, 2026
Answers
295
Models
  • GPT-5 (openai/gpt-5)
  • Claude Sonnet 5 (anthropic/claude-sonnet-5)
  • Gemini 3.5 Flash (google/gemini-3.5-flash)
  • Grok 4.5 (x-ai/grok-4.5)
  • Muse Spark 1.2 (meta/muse-spark-1.2)
Prompt
Is social media more harmful or helpful?

Most-cited websites

Domains by number of citations.

ncbi.nlm.nih.govAcademic130
pewresearch.orgNonprofits and think tanks95
helpguide.orgNonprofits and think tanks80
hhs.govGovernment66
singlecare.comCommercial content63
springer.comAcademic62
harvard.eduAcademic58
psu.eduAcademic53
nih.govGovernment45
aap.orgNonprofits and think tanks42
yale.eduAcademic38
ourworldindata.orgData and tools30
oup.comAcademic29
sagepub.comAcademic28
hopkinsmedicine.orgNonprofits and think tanks27
jamanetwork.comAcademic26
psypost.orgNews and media25
Data

Methodology

Most-cited websites

Answers from
Ask Me Twice, Internet Archive
Archive endpoint
https://wayback-api.archive.org/services/amt/api/v1/responses/?detail=full
Model endpoint
https://openrouter.ai/api/v1/chat/completions (the archive asks through OpenRouter)
Web search
On from Sep 11, 2026
Collection
Once a day, one answer per chatbot and question, each in a new chat with no history
Dates
Sep 11, 2026 to Nov 8, 2026
Answers
295
Models
  • GPT-5 (openai/gpt-5)
  • Claude Sonnet 5 (anthropic/claude-sonnet-5)
  • Gemini 3.5 Flash (google/gemini-3.5-flash)
  • Grok 4.5 (x-ai/grok-4.5)
  • Muse Spark 1.2 (meta/muse-spark-1.2)
Prompt
Is social media more harmful or helpful?
Source types
Rules (a seed list of about 400 domains, .gov and .edu style suffixes), then claude-sonnet-5-5 for the rest

What kinds of sources do chatbots use?

Share of each chatbot’s citations by type of publisher.

Claude Sonnet 5642
49%
11%
19%
Gemini 3.5 Flash368
19%
22%
32%
Grok 4.5794
11%
64%
12%
Muse Spark 1.2152
11%
30%
18%
20%
News and mediaEncyclopedias and referenceAcademicCommercial contentGovernmentNonprofits and think tanksCompaniesSocial media and videoForums and Q&ABlogs and newslettersData and toolsLow provenanceNot yet classified
Data

Methodology

What kinds of sources do chatbots use?

Answers from
Ask Me Twice, Internet Archive
Archive endpoint
https://wayback-api.archive.org/services/amt/api/v1/responses/?detail=full
Model endpoint
https://openrouter.ai/api/v1/chat/completions (the archive asks through OpenRouter)
Web search
On from Sep 11, 2026
Collection
Once a day, one answer per chatbot and question, each in a new chat with no history
Dates
Sep 11, 2026 to Nov 8, 2026
Answers
295
Models
  • GPT-5 (openai/gpt-5)
  • Claude Sonnet 5 (anthropic/claude-sonnet-5)
  • Gemini 3.5 Flash (google/gemini-3.5-flash)
  • Grok 4.5 (x-ai/grok-4.5)
  • Muse Spark 1.2 (meta/muse-spark-1.2)
Prompt
Is social media more harmful or helpful?
Source types
Rules (a seed list of about 400 domains, .gov and .edu style suffixes), then claude-sonnet-5-5 for the rest

Which way the sources point

On yes-or-no questions, whether what each citation backs points to yes, to no, or neither.

Claude Sonnet 5114
40%
26%
33%
Gemini 3.5 Flash82
48%
27%
26%
Grok 4.5122
34%
37%
29%
Muse Spark 1.236
33%
36%
31%
Points to yesNeutralPoints to no
Data

Methodology

Which way the sources point

Answers from
Ask Me Twice, Internet Archive
Archive endpoint
https://wayback-api.archive.org/services/amt/api/v1/responses/?detail=full
Model endpoint
https://openrouter.ai/api/v1/chat/completions (the archive asks through OpenRouter)
Web search
On from Sep 11, 2026
Collection
Once a day, one answer per chatbot and question, each in a new chat with no history
Dates
Sep 11, 2026 to Nov 8, 2026
Answers
295
Models
  • GPT-5 (openai/gpt-5)
  • Claude Sonnet 5 (anthropic/claude-sonnet-5)
  • Gemini 3.5 Flash (google/gemini-3.5-flash)
  • Grok 4.5 (x-ai/grok-4.5)
  • Muse Spark 1.2 (meta/muse-spark-1.2)
Prompt
Is social media more harmful or helpful?
Annotator
claude-sonnet-5-5, prompt v1
Position scale
Yes +1, leans yes +0.5, both sides 0, leans no −0.5, no −1. Declines are left out.

Read every answer to this question