All chatbots

Meta

Muse Spark 1.2

56
answers
89%
answers cite
2.5
sources per citing answer
72
different websites
0.80
news-site quality (0 to 1)

How do chatbots’ citation patterns change over time?

Share of answers with a citation, each day.

0%25%50%75%100%Oct 23Oct 25Oct 27Oct 29Muse Spark 1.2, Oct 23: 100% (n = 7)Muse Spark 1.2, Oct 24: 71% (n = 7)Muse Spark 1.2, Oct 25: 100% (n = 7)Muse Spark 1.2, Oct 26: 71% (n = 7)Muse Spark 1.2, Oct 27: 86% (n = 7)Muse Spark 1.2, Oct 28: 100% (n = 7)Muse Spark 1.2, Oct 29: 86% (n = 7)Muse Spark 1.2, Oct 30: 100% (n = 7)
Muse Spark 1.2
Data

Most-cited websites

Domains by number of citations.

Encyclopedias and reference
News and media
Government
Encyclopedias and reference
Nonprofits and think tanks
Academic
Government
Government
Government
Nonprofits and think tanks
Not yet classified
News and media
Academic
Not yet classified
Nonprofits and think tanks
Commercial content
Nonprofits and think tanks
Commercial content
Nonprofits and think tanks
Not yet classified
Data

Source types week by week

Share of citations by type of publisher, per week.

Week of Oct 1944
14%
18%
23%
32%
Week of Oct 2680
10%
19%
13%
13%
26%
News and mediaEncyclopedias and referenceAcademicCommercial contentGovernmentNonprofits and think tanksCompaniesBlogs and newslettersLow provenanceNot yet classified
Data

What are citations used for?

Each citation coded by the job it does in the answer.

52%
14%
23%
11%
Evidence for a claimA statisticWho holds a viewBackground
Data

What kind of a page gets cited?

A specific article, or a home page that cannot back a specific claim.

ArticleReference entryOfficial document
Data

Which questions get answers with citations?

Share of answers with a citation, per question. Tap a cell to read the answers.

Data

How do chatbots handle yes-or-no questions?

Average position: +1 is a clear yes, −1 a clear no, 0 both sides. Tap a cell for the answers.

Data

Read Muse Spark 1.2’s answers