Bluesky Cluster Analysis Report - August

📋 Table of Contents

📊 Overview & Analysis

🎯 Detailed Sections

📈 Additional Analysis

💡 Navigation Tip: Click any section title to jump directly to that part of the report. Use your browser's "Find" function (Ctrl/Cmd+F) to search for specific terms.

📋 Methodology & Data Collection

Report Generated: 2025-09-15 18:34:07

Session: august

Date Range: 2025-08-01 to 2025-08-31

Search Keywords (36 total):

aigrok#aichatgptchatbotopenaigeminillmclaudexaillmsgptartificial intelligenceanthropicgenaichatbots#genai#chatgpt#artificialintelligencenvidiacopilotmachine learninglanguage modellanguage modelsaltmantraining datadata centerdata centersrobotrobotsagideepmindalignment problemdeep learningneural networkprompt engineering

Data Processing Pipeline:

  1. Data Collection: 679,509 posts downloaded via Bluesky API using above keywords
  2. Like Threshold: Filtered to posts with ≥10 likes
  3. Content Filtering: Removed posts labeled as:
    • Spam content
    • Adult content (porn, sexual, nudity, sexual-figurative, graphic-media)
  4. Reply Enrichment: Added parent/root post context for replies
  5. Semantic Extraction: Generated semantic content for embedding
  6. Clustering: Applied UMAP + Gaussian Mixture Model (GMM) clustering algorithm
  7. AI Classification: Cluster descriptions and relevance labels generated by GPT-5 mini

Final Dataset Statistics:

↑ Back to top

📊 Executive Summary

57,549
Relevant Posts
38
Relevant Clusters
4,408,373
Total Likes
76.6
Avg Likes/Post
24,862
Unique Authors

📍 2D Cluster Visualization

Interactive scatter plot showing the spatial distribution of posts from relevant clusters in 2D embedding space. Each point represents a post with ≥100 likes, colored by cluster assignment.

Plotting 6,780 high-engagement posts (≥100 likes) across 38 clusters

↑ Back to top

📋 Relevant Clusters Summary

Overview of the 38 relevant clusters identified in the analysis, ordered by post count.

Cluster ID Posts Total Likes Avg Likes/Post Description
Cluster 1 5,871 349,154 59.5 AI Excuse Shaming
Cluster 2 4,985 329,441 66.1 Robots Replace Workers
Cluster 3 3,237 167,898 51.9 Debunking LLM Hype
Cluster 4 3,013 308,044 102.2 Moralizing ChatGPT Use
Cluster 5 2,814 184,990 65.7 Cultural AI Alarm
Cluster 6 2,609 152,485 58.4 Classroom AI Backlash
Cluster 7 2,177 146,003 67.1 Resisting AI Slop
Cluster 8 2,012 164,299 81.7 GPT Hype Backlash
Cluster 9 1,926 154,263 80.1 AI Film Culture Clash
Cluster 10 1,688 170,238 100.9 AI Bubble Backlash
Cluster 11 1,624 122,124 75.2 AI Surveillance Protests
Cluster 12 1,555 109,584 70.5 AI Panic Politics
Cluster 13 1,534 188,065 122.6 Grok Drama Cycle
Cluster 14 1,493 135,376 90.7 Deepfakes & AI Slop
Cluster 15 1,437 83,313 58.0 YouTube AI Backlash
Cluster 16 1,431 158,514 110.8 ChatGPT Suicide Fallout
Cluster 17 1,365 97,668 71.6 Toxic AI Aesthetics
Cluster 18 1,349 146,552 108.6 Altman Roast Chorus
Cluster 19 1,336 86,611 64.8 AI Slop Backlash
Cluster 20 1,116 105,396 94.4 Disable AI Features
Cluster 21 1,109 78,408 70.7 Art vs AI Activism
Cluster 22 1,082 87,427 80.8 AI Energy Backlash
Cluster 23 1,033 58,051 56.2 AI's Climate Footprint
Cluster 24 980 58,094 59.3 Authors vs AI
Cluster 25 957 85,667 89.5 AI Romance Panic
Cluster 26 930 64,099 68.9 Training Data Transparency
Cluster 27 903 71,832 79.5 AI Bubble Politics
Cluster 28 826 75,632 91.6 AI Harms & Ethics
Cluster 29 789 55,885 70.8 AI Surveillance and Harms
Cluster 30 705 83,573 118.5 Microsoft Copilot Backlash
Cluster 31 597 59,758 100.1 Dangerous AI Advice
Cluster 32 573 41,118 71.8 AI and the Dead
Cluster 33 570 44,290 77.7 Medical AI Alarm
Cluster 34 515 67,351 130.8 Meta Chatbot Scandal
Cluster 35 494 43,111 87.3 Authors vs Anthropic
Cluster 36 490 46,387 94.7 Platform AI Backlash
Cluster 37 266 22,953 86.3 Gemini Meltdown Memes
Cluster 38 158 4,719 29.9 Big Brother AI Backlash
↑ Back to top

🎯 Individual Cluster Analysis

Detailed analysis of each relevant cluster, ordered by number of posts.

Cluster 1

5,871
Posts
349,154
Total Likes
59.5
Avg Likes/Post
4,430
Unique Authors

📝 AI Excuse Shaming

Viscerally anti‑AI, gatekeeping discourse from artists and critics who angrily defend craft — these posts attack people who use generative AI as a "crutch," emphasize that art is learned work, and call out models as built on "stolen" creative labor. The cluster is cultural/ethical and emotional rather than technical or news-focused, often profane and illustrated with comics/art.

🏷️ Keywords

I don't use aijust a toolstolen workneed AIexcuses

📸 Top 3 Representative Posts

Post by 🌙Sonny Ross🌞
#1: 🌙Sonny Ross🌞
View original
Post by merritt
#2: merritt
View original
Post by Littleevil0ne *COMMS OPEN*
#3: Littleevil0ne *COMMS OPEN*
View original

🏆 Top 5 Authors in this Cluster

Rank Author Posts Total Likes Avg Likes/Post
1 🌙Sonny Ross🌞
📊 Top 8 Overall | @sonnyross.bsky.social		 8 23,230 2903.8
2 merritt
📊 Top 16 Overall | @merrittk.com		 3 8,687 2895.7
3 Littleevil0ne *COMMS OPEN*
@littleevil0ne.bsky.social		 7 5,652 807.4
4 Maddie Jacobus
@tacobusart.bsky.social		 5 4,889 977.8
5 alix e. harrow
@alixeharrow.bsky.social		 1 4,368 4368.0

Cluster 2

4,985
Posts
329,441
Total Likes
66.1
Avg Likes/Post
3,277
Unique Authors

📝 Robots Replace Workers

A cohesive, labor-oriented critique of AI that ties generative models and automation to economic injustice: job loss, wage suppression, corporate enrichment, and harm to creators and society. Unlike the comparison posts (practical tips, varied tech news, or neutral commentary), this cluster is explicitly political — pro-union, anti-oligarch, and doom-tinged about AI’s social impact.

🏷️ Keywords

replace you with a robotimpoverishes ussever capital from laborwage theftdepress wages

📸 Top 3 Representative Posts

Post by Mister
#1: Mister
View original
Post by Dan Johnson
#2: Dan Johnson
View original
Post by Charlie Warzel
#3: Charlie Warzel
View original

🏆 Top 5 Authors in this Cluster

Rank Author Posts Total Likes Avg Likes/Post
1 Ed Zitron
📊 Top 1 Overall | @edzitron.com		 23 14,373 624.9
2 Mister
@2snarky.bsky.social		 1 8,833 8833.0
3 Dan Johnson
@coil.bsky.social		 3 7,428 2476.0
4 Faine Greenwood
📊 Top 2 Overall | @faineg.bsky.social		 23 6,659 289.5
5 Charlie Warzel
@cwarzel.bsky.social		 4 5,992 1498.0

Cluster 3

3,237
Posts
167,898
Total Likes
51.9
Avg Likes/Post
1,895
Unique Authors

📝 Debunking LLM Hype

A skeptical, critical cluster focused on LLM limitations and hype — calling out hallucinations, the purely statistical nature of models, and how fawning conversational design tricks people into attributing understanding. These posts emphasize epistemic caution, debunking anthropomorphism, and situating current AI chatter in historical/ethical critique.

🏷️ Keywords

LLMshallucinationstatistical enginesfawning conversational tonenot search engines

📸 Top 3 Representative Posts

Post by Mike Masnick
#1: Mike Masnick
View original
Post by Robert Evans (the Only Robert Evans)
#2: Robert Evans (the Only Robert Evans)
View original
Post by stellz
#3: stellz
View original

🏆 Top 5 Authors in this Cluster

Rank Author Posts Total Likes Avg Likes/Post
1 Faine Greenwood
📊 Top 2 Overall | @faineg.bsky.social		 45 10,952 243.4
2 Mike Masnick
@mmasnick.bsky.social		 5 4,920 984.0
3 Robert Evans (the Only Robert Evans)
@iwriteok.bsky.social		 1 3,143 3143.0
4 Will Stancil
@whstancil.bsky.social		 48 3,003 62.6
5 stellz
@piss.beauty		 3 2,743 914.3

Cluster 4

3,013
Posts
308,044
Total Likes
102.2
Avg Likes/Post
2,358
Unique Authors

📝 Moralizing ChatGPT Use

This cluster is defined by moral and social condemnation of people who use ChatGPT—framing usage as embarrassing, a loss of creativity/agency, and a form of deskilling or grooming. Posts emphasize psychosocial harms (loneliness, sycophancy, delusion, cult-like adoption) and mock users, rather than focusing on technical, policy, or industry-level critiques seen in other clusters.

🏷️ Keywords

ceding your mental and creative abilitiesdeskilling is realgroominglonelier than everembarrassing

📸 Top 3 Representative Posts

Post by Ursula Dorada
#1: Ursula Dorada
View original
Post by Boze the Library Owl
#2: Boze the Library Owl
View original
Post by Trung Le Nguyen
#3: Trung Le Nguyen
View original

🏆 Top 5 Authors in this Cluster

Rank Author Posts Total Likes Avg Likes/Post
1 Ursula Dorada
📊 Top 5 Overall | @sulamoon.bsky.social		 1 24,803 24803.0
2 Trung Le Nguyen
📊 Top 11 Overall | @trungles.com		 5 15,078 3015.6
3 Boze the Library Owl
@sketchesbyboze.bsky.social		 1 14,587 14587.0
4 Andrew Lawrence
@ndrew.bsky.social		 3 12,773 4257.7
5 Chris Person
@papapishu.bsky.social		 2 9,298 4649.0

Cluster 5

2,814
Posts
184,990
Total Likes
65.7
Avg Likes/Post
2,048
Unique Authors

📝 Cultural AI Alarm

A cluster of opinionated, culturally-focused backlash against AI — skeptical essays, comics, and metaphors arguing that AI’s social, political, ecological, and psychological harms are being normalized. Distinctive for literary/metaphoric language (Asbestos, Butlerian Jihad, fae), anti‑anthropomorphism calls, and activism-style critique of AI’s institutional spread.

🏷️ Keywords

pretending they are humanAsbestos as a metaphorButlerian Jihadecological costschatbots are essentially the fae

📸 Top 3 Representative Posts

Post by Smooth Dunk
#1: Smooth Dunk
View original
Post by Hannah Bluish
#2: Hannah Bluish
View original
Post by Kat Tenbarge
#3: Kat Tenbarge
View original

🏆 Top 5 Authors in this Cluster

Rank Author Posts Total Likes Avg Likes/Post
1 Wendy Xu
📊 Top 18 Overall | @wendyxu.bsky.social		 2 9,461 4730.5
2 Faine Greenwood
📊 Top 2 Overall | @faineg.bsky.social		 21 6,724 320.2
3 Smooth Dunk
📊 Top 7 Overall | @smoothdunk2.bsky.social		 1 4,954 4954.0
4 Kat Tenbarge
📊 Top 3 Overall | @kattenbarge.bsky.social		 6 4,938 823.0
5 Hannah Bluish
@hannahbluish.bsky.social		 1 4,566 4566.0

Cluster 6

2,609
Posts
152,485
Total Likes
58.4
Avg Likes/Post
1,611
Unique Authors

📝 Classroom AI Backlash

This cluster is tightly focused on education: negative student and teacher reactions to AI, concerns about academic integrity, and arguments that AI adoption erodes critical thinking and creative pedagogy. Posts emphasize resistance (avoidance, hatred), examples of cheating and administrative pressure, and a socioeconomic critique of elite vs working‑class schooling.

🏷️ Keywords

hate AIavoid AI chat botscritical-thinkingcheating with ChatGPTAI in education

📸 Top 3 Representative Posts

Post by Benjamin Balthaser
#1: Benjamin Balthaser
View original
Post by Will Mason
#2: Will Mason
View original
Post by Melissa Johnson, PhD
#3: Melissa Johnson, PhD
View original

🏆 Top 5 Authors in this Cluster

Rank Author Posts Total Likes Avg Likes/Post
1 Benjamin Balthaser
@blbalthaser.bsky.social		 2 4,098 2049.0
2 Faine Greenwood
📊 Top 2 Overall | @faineg.bsky.social		 4 3,553 888.2
3 Seth Cotlar
@sethcotlar.bsky.social		 12 3,248 270.7
4 Melissa Johnson, PhD
@ladyhistorian.bsky.social		 15 3,238 215.9
5 Aparna Nair
@disabilitystor1.bsky.social		 35 3,121 89.2

Cluster 7

2,177
Posts
146,003
Total Likes
67.1
Avg Likes/Post
1,675
Unique Authors

📝 Resisting AI Slop

A strongly skeptical, often angry cluster focused on AI's cultural and labor harms — calling out training‑data theft, creator burnout, corporate bad faith, and the spread of low‑quality ‘AI slop’ and misinformation that contaminates sources like Wikipedia. These posts defend human work, distrust model outputs, and emphasize real-world consequences (education, authorship, art).

🏷️ Keywords

AI slopstole training databurnt outNot ChatGPT outputWikipedia

📸 Top 3 Representative Posts