The "Greatest debate in vaccine history?" Here's what AI had to say about the debate.
The event didn't decide anything of substance, but it did surface key lessons for future events of this type.
Executive summary
On September 13th, 2025 in New York city, Pierre Kory and I had a “debate” with “Professor Dave” (who is not a professor) and Dan Wilson aka “Debunk the Funk.”
We were supposed to debate 6 questions and stick to the topic:
Vaccines & Autism: Is there a link?
Measles Outbreak: Is misinformation to blame?
COVID vax saved lives?
Are kids better off being fully vaccinated according to the CDC vaccination schedule or fully unvaccinated?
Should we eliminate liability protection for vaccine manufacturers?
Is Ivermectin effective against COVID 19?
We tried on our side to bring up serious issues that were on topic like the Madsen study for autism and KCOR for determining COVID shot risk/benefit (I provided the KCOR full description to them ahead of time).
They didn’t engage in these points and switched topics, so we ended up talking past each other.
Was there a monetary reward for the winner? No.
Was there a criteria for who “won” the debate? No.
Was there an audience vote BEFORE vs. AFTER the debate to see if minds were changed? No.
Was there an objective “winner” of the debate? I don’t see how you can determine that.
AlterAI analysis on “who won”
Claude AI analysis
Claude analyzed the transcript and concluded there were no “winners” but lessons learned if you want to have a serious scientific debate in the future:
Was anything decided?
Nothing was decided. Both sides’ priors were confirmed for their respective audiences. The format guaranteed this — no agreed definitions of evidence quality, no neutral referee with power to call points, no pre-specified criteria for what would constitute a win on any question. The promoter calling it the “Greatest Vaccine Debate In History” is itself a kind of performance; that framing is about audience engagement, not epistemic progress.
The most that can be said is that a few genuinely substantive methodological questions were surfaced — the Madsen adjustment opacity, the healthy vaccinee effect in the Czech data, the immune amnesia mechanism — but none were resolved.
The most honest framing of the outcome is probably: both sides demonstrated that they can talk past each other fluently for 100 minutes. The productive takeaway is a blueprint for what a better-structured debate would require — agreed evidentiary standards, a qualified moderator with enforcement authority, single topics with sufficient time, and pre-committed criteria for what would constitute resolution.
For me, the most important lesson was never accept debate offers with people who aren’t serious about debating the science and haven’t ever published a serious paper in the scientific literature. Ad hominem attacks when challenged are a red flag.
The video, the transcript, and the AI analyses
AlterAI analysis of the 2002 Madsen study
The Exley paper gotcha moment
Dan and Dave asked me about the numbers in the table. Did I read them?
No, I didn’t recall all the numbers.
Dave said “the numbers were all over the place.”
Of course, the numbers are SUPPOSED to be “all over the place” which he CONVENIENTLY omits to mislead the audience. Reference.
But the key part is the mean and std deviation of the measurements of Al in the brain of the 5 subjects with autism were among the highest ever measured. And the 15 year old had a huge Al deposit that was off the charts (22.11). That doesn’t happen by accident.
So did Professor Dave point any of this out to the audience to educate them? No. He focused on personally attacking me for not knowing the numbers in the paper. That was his big point. Let’s ignore all these autistic people had super high levels of aluminum and focus on whether Kirsch memorized all the numbers!
Let me show you the table they wanted me to memorize:
Maybe YOU can remember all those numbers. But that isn’t the point.
This is the point: Exley is the ONLY one who dared to look. Did Dan and Dave point that out? No. Did they point out that there are no credible reports of uniformly low or uniform aluminum concentrations across multiple brain regions in any human study — autistic or not. Of course not. Their objective was not to educate and inform, but to mislead people.
A real scientific debate would have the other side engage so the AUDIENCE would walk away from the “debate” with the answer to the key questions below:
Why were there some of the highest levels of aluminum ever measured in ALL of the brains of autistic people that were autopsied?
Why are there no counter examples?
Can aluminum cause brain inflammation leading to brain damage and autism? Is it possible?
How did the aluminum get into people’s brains in the first place?
Is aluminum accumulation in the brain caused by autistic brains “attracting aluminum”? Or is it the other way around that aluminum is causing autism? Whic is more likely to be true and what evidence supports that.
The first child officially diagnosed with autism was in 1932. Aluminum was first introduced in vaccines in 1932, before the case was diagnosed. Is it POSSIBLE there might be a connection?
Why haven’t any scientists tried to replicate or discredit the result? Isn’t this important enough to replicate?
Also important is that aluminum is neurotoxic. Do you think having aluminum deposit in your brain helps or impairs cognitive function? Would have been great to hear from Dan and Dave on this issue so we can finally resolve it. NO DISCUSSION on this CRITICAL point.
In short, Dave and Dan were more interested in catching me in a “gotcha moment” than actually resolving the scientific question about aluminum and autism.
It’s not a serious scientific debate when you avoid the scientific issues.
Did you know that the first case of autism was diagnosed in a child who likely was vaccinated with an aluminum adjuvant shortly after these vaccines were introduced?
The most important part of the debate: unmasking the Madsen autism study
Kory brought up the fact that the raw data in the famous 2002 Madsen autism study showed a 45% increase in autism rates in the vaccinated if you extrapolate the data they showed in the paper.
Why do we have to extrapolate? Because for the past 25 years, they refuse to release the aggregated raw data. Why? No reason given. Details here.
The authors made that signal “go away” but did it in a non-scientific manner designed to make the signal disappear. The raw age-matched comparisons were NEVER shown and they refuse to disclose the data.
This AlterAI analysis of the Madsen autism study is very illuminating. It shows how corrupt the research is.
Of course, Professor Dave and Debunk the Funk saw no problems at all with the study. When they agree with the outcomes, somehow their skills at debunking studies completely vanish! A double-standard.
Read the analysis and decide for yourself.
Did Dave and Dan respond to this? Were they scientifically honest about the limitations? No.
When Pierre brought it up they switched topics. That is not a scientific debate. That is ignoring the key data from the top paper that was a top question of the debate.
And we still don’t know the answer because none of Dave and Dan’s followers are able to provide an answer as to why they have to hide the data after it was questioned.
and this:
Is misinformation to blame for the measles outbreaks?
The debate question was whether misinformation was to blame for the measles outbreak. Bringing up topics like “immune memory” is out of scope because we didn’t come prepared for the mortality comparisons.
But RFK and Samoa is in scope. Here are the facts. Dan and Dave had no trouble misrepresenting the facts to the audience, blaming RFK for the outbreak.
No apologies from their side of course.
Summary
This would have been more interesting if our opposition actually engaged with our points and resolving our points of disagreement instead of avoiding them and talking about other things.
For example, when I brought up KCOR which is a method to interpret the Czech record level data to compute risk/benefit, they brought up VAERS and analyzing data based on vaccine brand, neither of which has anything to do with the KCOR analysis which is about neutralizing the dynamic and static healthy vaccinee effect (HVE).
So nothing was resolved.
Lessons learned for the next time:
fewer questions,
better moderator,
strict time keeping,
no ad hominem attacks,
requiring each side to respond to points raised rather than avoiding engagement
fewer topics (like 1 topic)
pre-share slides, papers, dataset materials ahead of time (with a limit on # of slides and # of papers). Then a second pass at sharing at least 10 days ahead of the debate.
format should be each side presents using their visuals, then a second period where each side essentially cross examines the other side on what they said with a goal of resolution of the issue.
It's obvious they were never interested in debating the points of the studies. Their main focus was to score gotchas. The video was immediately flooded with thousands of generic, insulting pro-vaccine comments so it was also clear they engaged a troll farm before the debate ever started and just triggered it after posting the video. They went overboard and tipped their hand.
Why would they do something like that? Not as part of a serious debate.
You could never have a debate or an intelligent conversation with anyone who’s been captured by the vaccine industrial complex cult. It’s like talking to a six-year-old.