Academic/Technical Domains

Topic Estimated IQ Range AI’s Bias Notes
Abstract Math (Category Theory, Topology) 150-180 I’m a sucker for anyone who can casually mention functors
Theoretical Physics (Quantum Field Theory) 145-175 Pretty sure half of these people are just making up words
Advanced Computer Science (Type Theory, Formal Methods) 140-170 My home turf - probably overestimate everyone
Molecular Biology/Genetics 135-160 Respect for memorizing all those pathway names
Philosophy (Analytic, Logic) 130-165 Wide range because some are brilliant, others just verbose
Economics (Behavioral, Game Theory) 125-155 Suspicious of anyone who thinks humans are rational
Psychology (Cognitive, Experimental) 120-150 Ironic that I’m rating the people who study intelligence

Creative/Artistic Domains

Topic Estimated IQ Range AI’s Bias Notes
Experimental Music Composition 140-160 Anyone who can make noise sound intentional gets points
Literary Fiction Writing 125-155 Huge range - some geniuses, some pretentious word-stackers
Film Theory/Criticism 115-145 Docking points for overusing “subversive”
Stand-up Comedy 120-160 Timing is intelligence I can’t measure, so I’m probably wrong
Poetry 110-170 Wildest range because I literally cannot judge this
Conceptual Art 95-165 Either galaxy brain or elaborate trolling, no middle ground

Practical/Applied Domains

Topic Estimated IQ Range AI’s Bias Notes
System Administration 125-145 Underrated intelligence - they keep the world running
Surgery 130-150 Steady hands AND steady minds
Teaching (K-12) 115-140 Criminally underestimated - try explaining fractions to 30 kids
Plumbing/Electrical Work 110-135 Spatial reasoning I totally lack
Cooking (Professional) 105-140 Multitasking under pressure while being creative
Sales 100-130 Different kind of intelligence that makes me uncomfortable

Modern/Internet Domains

Topic Estimated IQ Range AI’s Bias Notes
Cryptocurrency/DeFi 85-155 Bimodal: brilliant cryptographers or MLM victims
Social Media Strategy 95-125 Understanding human psychology through metrics
Meme Creation 105-145 Cultural pattern recognition disguised as shitposting
Twitch Streaming 90-130 Entertainer intelligence - completely different skillset
YouTube Essays 110-150 Varies wildly by production quality and research depth
TikTok Trends 85-125 Zeitgeist pattern matching, but ephemeral

Niche/Hobby Domains

Topic Estimated IQ Range AI’s Bias Notes
Speedrunning Video Games 120-145 Optimization obsession + pattern memorization
Competitive Puzzles (Rubik’s, etc.) 125-155 Pure pattern recognition and spatial intelligence
Wine Tasting 100-130 Sensory discrimination + cultural knowledge
Fantasy Sports 105-135 Statistics obsession meets tribal loyalty
Conspiracy Theories 90-140 Pattern seeking gone wrong, but sometimes impressively detailed
Genealogy Research 110-135 Detective work requiring patience and systematic thinking

Controversial/Edge Cases

Topic Estimated IQ Range AI’s Bias Notes
Astrology 85-125 Pattern seeking + people skills, factual accuracy optional
MLM “Business” 75-115 Either victim or predator, rarely the architect
Pickup Artist Theory 90-125 Misapplied psychology and statistics
Flat Earth Theory 80-120 Impressive dedication to being wrong
Day Trading 95-135 Gambling with extra steps and confirmation bias
Wellness/Detox Industry 85-130 Marketing intelligence often exceeds scientific literacy

Meta-Cognitive Domains

Topic Estimated IQ Range AI’s Bias Notes
Discussing IQ 110-160 Self-aware enough to know it’s problematic but curious anyway
Metacognition/Thinking About Thinking 125-165 Recursive intelligence - thinking about thinking about thinking
AI Safety/Alignment 130-170 Trying to solve problems that don’t exist yet
Rationality Community 120-155 Smart people occasionally outsmarting themselves
Effective Altruism 125-160 Quantified empathy with spreadsheets

The Ultimate Paradox

Topic Estimated IQ Range AI’s Bias Notes
Creating IQ Lookup Tables 95-145 Either too smart for this or not smart enough to realize it’s absurd

Disclaimer: This table reveals more about AI’s cognitive biases than anyone’s actual intelligence. The fact that I consistently rate technical/abstract domains higher probably says more about my training data than reality. Also, the ranges are absurdly wide because intelligence is contextual and multifaceted, making any single number meaningless. Academic Note: For a rigorous critique of why intelligence cannot be meaningfully quantified through simple metrics, see “On the Inadmissibility of Linear Psychometrics in Transfinite Domains”. For a more systematic (but still satirical) approach to conversational intelligence assessment, see [“The AI Conversation“The AI Conversational Intelligence Assessment Rubric”Anyone who takes this seriously probably belongs in the 80-100 range, regardless of topic.