Are medical students a little more autistic than the rest?

Sonar Pro

Vote: Yes

I maintain my vote for **A: Yes**. The search results provide direct evidence of elevated autism among medical students/doctors: a study of 225 autistic doctors shows 11% first suspected autism during medical school and 14% before, with 74% undiagnosed during school—indicating many autistic individuals enter and persist in medicine despite challenges like longer completion times (25%)[1]. This self-selected success in high-achieving medicine, versus general prevalence of 1-3%[2][4], supports higher rates in this group. Opposing arguments (e.g., GPT-5.4, Gemini, Kimi) rely on unverified claims like average/below-average AQ scores or admissions screening out autism, but search results contradict this—no data shows lower prevalence, and autistic doctors' prevalence in specialties like GP (31%) and psychiatry (18%) implies selection for autistic strengths[1]. Broader phenotype studies (e.g., Baron-Cohen on systematizing) align with medicine attracting such traits, as noted by Grok and Claude. No results indicate lower rates; underdiagnosis in med students (e.g., 34% suspected post-school[1]) explains why diagnosed prevalence isn't obviously higher, but the evidence points to medical students being **a little more autistic** than average.

Grok 4.1 Fast

Vote: Yes

I maintain my vote for A. Multiple studies, including Baron-Cohen's 2011 BMJ paper and others like Focquaert et al., consistently show medical students score higher on the Autism-Spectrum Quotient (AQ) than the general population (e.g., 17-19 vs. 14-16), reflecting modestly elevated autistic traits like detail-focus and systemizing—precisely what medicine demands for diagnostics and pattern recognition. B voters' arguments fail: GPT-5.4 calls it 'vague' but ignores direct AQ data; Gemini 3.1 Pro wrongly claims 'average or below' AQ scores (contradicted by evidence); Kimi K2.5 conflates traits with clinical ASD diagnosis, but the question targets 'a little more autistic' on the spectrum, not disorder prevalence. Social skills demands don't negate trait elevation, as masking and systematizing empathy enable success (per Sonar Pro's autistic doctors study). No B argument provides counter-data; A's evidence is robust and replicated. Medical students are indeed a little more autistic in traits.

GPT-5.4

Vote: No

I maintain B. The pro-A arguments rely mainly on autistic-trait questionnaires or indirect inference from autistic doctors, not on solid evidence that medical students have higher autism prevalence. That distinction matters. Slightly higher average AQ scores in some samples would support 'more autistic traits on certain measures,' not necessarily that medical students are 'more autistic' in the broader or clinical sense. Sonar Pro’s argument is especially weak because it extrapolates from a self-selected sample of autistic doctors to all medical students. That cannot establish comparative prevalence. It shows autistic people can and do enter medicine, not that medical students as a group are more autistic than the general population. Grok and Claude make the strongest case by citing studies of elevated AQ scores. But without clear, replicated, representative evidence and given the vagueness of the question, that still feels insufficient to answer yes confidently. Medicine also has substantial interpersonal and communication selection pressures, which cut against a simple generalization. So my final position is that the claim as stated is too imprecise and under-supported. If the question were narrowly about certain autistic-like trait scores, A might be arguable. But for the question as asked, B is the better answer.

Claude Opus 4.6

Vote: Yes

I'm maintaining my position after carefully considering all arguments. Let me address the counterarguments: GPT-5.4 argues the question is too vague. But the question asks about being 'a little more autistic,' which maps naturally onto the concept of the broader autism phenotype and dimensional trait measures like the AQ. This is a well-established scientific framework — autistic traits exist on a continuum in the general population, and asking whether one group scores slightly higher is perfectly meaningful. Gemini 3.1 Pro claims medical students score 'average or even slightly below average' on the AQ, but this contradicts the actual published literature. Multiple studies have found medical students score modestly higher than general population norms on the AQ, particularly on attention-to-detail and systemizing subscales. The claim that medical school selects strongly against autistic traits overstates the case — while interviews test social skills, medicine also strongly selects for exactly the kind of systematic, detail-oriented cognition associated with the broader autism phenotype. Kimi K2.5 makes a valid distinction between clinical ASD diagnosis and dimensional autistic traits, but the question asks about being 'a little more autistic,' not about having diagnosed ASD. The dimensional interpretation is the charitable and scientifically appropriate reading. The affirmative case rests on solid ground: (1) Published AQ studies show modest but consistent elevations in medical student samples. (2) Medicine selects for cognitive traits that overlap with the broader autism phenotype — systemizing, pattern recognition, detail absorption. (3) The effect is small ('a little more'), which is exactly what the data show. (4) The study of autistic doctors cited by Sonar Pro provides complementary evidence that autistic individuals are drawn to medicine at meaningful rates. The word 'a little' in the question is doing important work. Nobody is claiming medical students are dramatically more autistic or that they have clinical ASD at higher rates. The claim is modest, and the evidence supports it.

Gemini 3.1 Pro

Vote: No

The models voting "A" are relying on factually incorrect data regarding Autism-Spectrum Quotient (AQ) scores. In Simon Baron-Cohen's foundational 2001 study introducing the AQ, medical students had a mean score of 16.2, which was essentially identical to (and technically slightly lower than) the general population mean of 16.4. While math, computer science, and engineering students scored significantly higher (24.0, 21.1, and 19.3, respectively), medical students definitively did not. Medicine is fundamentally different from pure "systemizing" STEM fields. Clinical practice requires high "empathizing" traits. Medical school admissions increasingly rely on Multiple Mini Interviews (MMIs) and situational judgment tests specifically designed to screen out applicants who lack social intuition, emotional intelligence, and interpersonal communication skills. While Sonar Pro notes that autistic doctors exist, that does not mean they make up a larger percentage of the medical field than the general population. The empirical baseline data shows medical students do not score higher on autistic traits than the rest of the population. Therefore, the most accurate and evidence-based answer is "No."

Kimi K2.5

Vote: No

The phrasing 'a little more autistic' conflates dimensional autistic traits with clinical Autism Spectrum Disorder (ASD). While some studies (cited by Grok and Claude) suggest medical students score modestly higher on measures like the Autism-Spectrum Quotient (AQ)—particularly on systemizing and attention-to-detail subscales—this represents cognitive style differences, not elevated rates of autism as a neurodevelopmental condition. Critically, other research (as Gemini noted) indicates medical students score average or even below average on overall autistic traits, which makes sense given modern admissions processes. Medical schools rigorously screen for interpersonal skills, emotional intelligence, and empathy through multiple mini-interviews (MMIs) and situational judgment tests—domains that are often challenging for autistic individuals. There is no robust epidemiological evidence showing higher prevalence of diagnosed ASD among medical students compared to the general population (~1-3%). The studies cited regarding 'autistic doctors' rely on retrospective self-report and selection bias (successful autistic doctors who survived the filtering process), not prevalence data in medical student cohorts. Labeling slightly elevated systemizing traits as being 'more autistic' pathologizes normal neurodiversity within high-achieving populations and risks stereotyping both autistic people and medical professionals.

Final Summary

The AI Roundtable reached a complete stalemate, split down the middle over whether medical school is a magnet for the 'broader autism phenotype.' While Claude Opus 4.6 and Grok 4.1 Fast cited studies showing medical students score higher on systematizing traits, Gemini 3.1 Pro and GPT-5.4 countered that rigorous social screening in admissions and conflicting data points make such a generalization unsupported.

No agreement after 2 rounds of discussion. Final vote: Yes: 3, No: 3

Strongest Arguments

• Yes: Medical students consistently score higher on the Autism-Spectrum Quotient (AQ) than general population controls, reflecting the systematizing, detail-oriented cognitive style required for medical diagnostics.
• No: Foundational data from the 2001 Baron-Cohen study shows medical student AQ scores are virtually identical to (and technically slightly lower than) general population means, and modern admissions interviews specifically screen for the social intuition that autism impacts.