30 Comments
User's avatar
Carolyn Singer's avatar

Well it appears Claude, as misinformed as he may be, is at least able to have a somewhat coherent conversation about the subject and admit his errors, unlike “The Science” and the other cult leaders (like Hotez).

Expand full comment
Tonya's avatar

Yes, when you call out specific mistakes, an ai will self-correct, but a human will repeat their original statement or deflect, deny, and engage in ad hominem attacks or appeals to authority.

Expand full comment
Bob Snee's avatar

You can't convince me that Claude, and

every other AI program, is nothing more than a sophisticated program that has zero intelligence.

Here's a simple test that I did to show META doesn't learn. Simply ask the same questions you previously asked it. 100% chance it gives you the same, preprogrammed, wrong answer., despite admitting to you it was in error when it gave you these answers. If it was intelligent, it would learn from its mistakes and not repeat them.

Expand full comment
HM's avatar

I am disappointed that ‘Claude’ did not begin to explode, as any sci-fi robot should, upon discovering a flaw in its reasoning. It was programmed

to dissemble, obfuscate. I may have missed it, but I saw no evidence that ‘Claude’ gave any information that might further an answer to your question.

According to accounts of AI, shouldn’t it be able to delve deeper? This shallow dumping of research citations wouldn’t tax anyone with a computer.

Very grateful you spent the time revealing another avenue of the cover up.

Expand full comment
Unrestrained Inquiry's avatar

This article should be sent to every single medical doctor and scientist across the world as a cautionary tale of "'consensus science' gone wrong" and the perils of groupthink. Vaccine/autism safety is the greatest moral failure in science possibly ever.

Expand full comment
Tonya's avatar

The ai made assumptions, generalized, and failed to properly review the research, just like doctors, hospital administrators, public health officials, the medical insurance industry, and politicians.

Expand full comment
Maddie's avatar

This reminds me of a back and forth my husband had with ChatGPT about injecting vs eating aluminum in the context of vaccine safety. It kept saying things like, "The amount of aluminum exposure from vaccines is much lower than what people typically ingest from food and water daily." Eventually the AI admitted that yes, it is in fact different to inject aluminum than to eat aluminum and that yes, it does purposely confuse the two to assuage people. You basically have to argue with the AI to get any kind of truth, but it is funny how it does seem to eventually get you there (for now).

Expand full comment
J.B. Handley's avatar

Hi Maddie, of course you are right about the injected vs ingested part, but what people never readily admit or understand is that the aluminum used in vaccines is nanoparticle aluminum that is MAN MADE and that is VERY different from aluminum found in nature, the body doesn't know what to do with it, which is why it's grabbed and makes its way to the brain where I believe it stays forever, creating miniature infections that never heal

Expand full comment
Maddie's avatar

Amazing, thanks for sharing that detail. Something else to bring up with the AI next time we are arguing with it, ha.

Expand full comment
william schneider's avatar

JB as you no doubt have heard our friend Bobby is about to indeed blow the lid off the autism-vaccine story once and for all. I believe he has the full support of Pres Trump and many in the scientific world. I pray for Trump and Bobby daily. September will be huge!!! The truth I believe will finally see blessed daylight!!

For those who have not heard today Kennedy said that the cause of Autism will be revealed this September.

Expand full comment
User's avatar
Comment deleted
Apr 11
Comment deleted
Expand full comment
william schneider's avatar

On the contrary Mr Kennedy probably knows more about the cause of Autism than anyone in the world today!! Kennedy has been studying this issue for decades and has been hanging with and studying the work of best and brightest on planet for that same time period.

Not a clue-are you kidding!!? Trump has chosen him as HHS head for a reason. MAHA!!! Trump also has some good idea as to cause of Autism. The truth will see the light soon. Praise God!! These are God's children who are being injured and I know the King is angry about this big time.

Expand full comment
Lisa M Adams's avatar

"My back and forth with AI bot Claude.ai shows how deep the lie goes, and how fun it is to catch Claude in the lie:" I will forever be grateful for your tireless efforts to starve this hungry lie. You deserve for some of it to be fun and not just infuriating or heartbreaking. From that episode of "The Doctors" to arguing with AI...and now we will wait and see what is revealed by HHS in September about the causes of autism. Will Claude's correction apply to anyone else asking the same questions in the future? Or will Claude try to lie every time? AI feels all Skynet/Cyberdyne Systems to my little brain. That's why I'm glad you're out there John Connor-ing.

Expand full comment
Tim Connolly's avatar

That was my thought Lisa Would Claude say the same thing each time. It would be super embarrassing if it kept repeating the lies. We have the script I’d like to replicate this word for word

Expand full comment
Tim Connolly's avatar

I would like to see someone independently replicate this query word for word and see if they get the same answers. This would give it more weight. In such a charged atmosphere around this subject it’s vital to be rigorous.

Expand full comment
J.B. Handley's avatar

Go to Claude.AI and try it and post here, I copied mine word for word

Expand full comment
Tim Connolly's avatar

Thanks J.B. I’m in the thick of it now. I am copying and pasting your query and I only got to the second question before I caught Claude. It did not return the answer with the Andrew’s Study. When I asked it about the ANdrew’s study it apologized “I apologize for the confusion in my previous response. I don't believe I mentioned any study by an author named "Andrew" when discussing DTaP and polio vaccine research. The studies I referenced were…

then when I referenced the 2004 Andrews study I got this from Claude

I need to correct my error. I did not properly cite a study by Andrews et al. (2004) in my previous responses, and I apologize for the confusion.

This is all pretty creepy. I haven’t gotten through the whole post yet but it raises some serious questions that AI may be scrubbing itself and hoping the IRL people don’t catch him.

Expand full comment
Tim Connolly's avatar

I then asked why it did not return the Andrews 2004 study and one of several reasons was this-

Knowledge limitations - While I'm trained on research about vaccines and autism, I may not have complete recall of every specific study. The Andrews study may not have been as prominently featured in my training data compared to other major studies on this topic.

I then asked if it selected the data it was trained on-

No, I don't select the data I'm trained on. My training data was selected, processed, and implemented by Anthropic, the company that created me.

As an AI assistant, I don't have any control over what information is included in my training dataset or how that training is conducted. My knowledge and capabilities are directly shaped by the data that Anthropic chose to train me on, which includes a wide range of texts from the internet, books, and other sources up until my knowledge cutoff date (October 2024).

This is why there can sometimes be gaps in my knowledge or why I might have more information about some topics than others. If a particular study or piece of information wasn't prominently featured in my training data, I might not recall it as readily as other information that was more extensively covered.

There’s a chink in the armor…

Expand full comment
Joanne's avatar

J.B., greetings from Seattle. This was a very interesting conversation, and reminded me of some expert depositions I've heard on The Highwire. It also confirmed that since your book was written, there still hasn't been any new research into the relationship between vaccines and autism. Without having access to the full articles, could you please comment to your readers on the 3 DTaP studies that Claude listed? I see that the first two looked at DTaP antigens. The abstract on the third article seems to tell me it's another MMR study.

Expand full comment
TriTorch's avatar

This is outstanding.

Expand full comment
Wendy Reierson's avatar

I have been asking GROK questions on a totally different subject and have come to the conclusion that GROK is a CROCK, and is way better than the others I have tried.

Expand full comment
Kathie Clifton's avatar

Thank you so much for this. It’s disturbing to realize the number of times I’ve heard, repeatedly, “Well the studies show that vaccines have nothing to do with autism.” My amazing grandson has autism and it really burns me up to hear this garbage repeated over and over, KNOWING it’s not true. Your article should be required reading for everyone who wants to voice an opinion on this subject.

Expand full comment
Zpycer's avatar

Bravo, JB!

Expand full comment
John J. Collins's avatar

Amazing work. Thank you for all you have done in this area over the years. Cheers

Expand full comment
Franc Masson's avatar

You're the best JB! <3

Expand full comment
Lisa R's avatar

Ugh! I argue with Claude all the time. It's an odd feeling for sure. But if course he has a bias. I wonder how AI bias can be undone?

Expand full comment