I don’t have a lot of experience at making judgments about orders of magnitude, so you are probably right. But my estimate was based not only on the seeming untrustworthiness of the source, but also on an estimate of how likely it is that anything like what this guy is claiming is true. If you multiply together the probabilities that:
literally all medical complaints, from autism to backache, have the same cause, which is deficiency of micronutrients available in ordinary foods if eaten in sufficient quantities;
the extremely vigorous biological/medical sciences have been fundamentally totally off the mark for at least a century and that probably something like hundreds of thousands of seemingly well-verified scientific findings are wrong (how this could have happened, the guy doesn’t say); and that
of all people, this clown[1] was the person who figured it out …
… I’m really not sure you get something much more likely than 1 in 10 million.
fn1. just an excuse to link to my favorite of his many hilarious comments (posted 8 1⁄2 hours after his previous announcement of a “breakthrough”):
Sweet another break through, was able to unify eastern and western medicine.
What are you trying to find by taking the product of these probabilities? They are not independent, so I think you meant to use probabilities that are conditional upon each other. For example, to find the probability of their conjunction, we would need
%20=%20P(A)%20\times%20P(B%20\mid%20A)%20\times%20P(C\mid%20B%20\cap%20A)) instead. 10^-7 seems a bit low for this.
Yeah, as I said to jsalvatier, it may well be that my estimate was too confident, and of course you’re quite right that I should have explicitly used conditional probabilities. We’re still dealing with someone who we can label a crackpot with only a trivial chance of being wrong.
There are currently 6 billion people on earth. In their lifetimes (presuming the singularity doesn’t extend that to infinity), I expect there will be significantly less than 600 fundamental theory changes to our systems. Sure, entirely new fields like computers and genetics have occurred in the last ~80ish years, but those occurred within the framework of existing theory. Few and far between are the refutations like Phlogiston, Impetus, or Spontaneous generation. The prior probability for anyone to overturn an established branch of science is therefore already much less than 1 in 10 million.
If our prior probability is already at 1 in 10 million, additional facts like him having no training in an area and incoherent sentences push it to being even less likely.
The relevant number to use in the calculation you’re making isn’t the 6 billion people inhabiting the world, but the people that actually claim to have made such a stunning breakthrough: since that’s what instigates us to want to evaluate the claim in the first place, the fact that he made it.
In short not:
P(X made a stunning breakthrough|X is an alive human being)
but rather:
P(X made a stunning breakthrough|X claims to have made a stunning breakthrough)
As a rough guess I’d say 1⁄100 (or atleast 1/1000) may be far closer in regards to prior probabilities than 1⁄10,000,000.
“We can not solve our problems with the same level of thinking that created them”. ― Albert Einstein
If modern scientific methods could self-explain why they were wrong, they would be better scientific models. Moreover, scientific communities do not have access to perfect knowledge: any particular theory could have hundreds of supporting trials behind it if those trials weren’t popular enough to be well-known (out of millions of experiments).
1e-7 seems too strong. That’s 1 in 10 million.
I don’t have a lot of experience at making judgments about orders of magnitude, so you are probably right. But my estimate was based not only on the seeming untrustworthiness of the source, but also on an estimate of how likely it is that anything like what this guy is claiming is true. If you multiply together the probabilities that:
literally all medical complaints, from autism to backache, have the same cause, which is deficiency of micronutrients available in ordinary foods if eaten in sufficient quantities;
the extremely vigorous biological/medical sciences have been fundamentally totally off the mark for at least a century and that probably something like hundreds of thousands of seemingly well-verified scientific findings are wrong (how this could have happened, the guy doesn’t say); and that
of all people, this clown[1] was the person who figured it out …
… I’m really not sure you get something much more likely than 1 in 10 million.
fn1. just an excuse to link to my favorite of his many hilarious comments (posted 8 1⁄2 hours after his previous announcement of a “breakthrough”):
Riiiiiiiiight.
What are you trying to find by taking the product of these probabilities? They are not independent, so I think you meant to use probabilities that are conditional upon each other. For example, to find the probability of their conjunction, we would need
%20=%20P(A)%20\times%20P(B%20\mid%20A)%20\times%20P(C\mid%20B%20\cap%20A))instead. 10^-7 seems a bit low for this.
That quote is great. :D
Yeah, as I said to jsalvatier, it may well be that my estimate was too confident, and of course you’re quite right that I should have explicitly used conditional probabilities. We’re still dealing with someone who we can label a crackpot with only a trivial chance of being wrong.
Actually, raw order of magnitude it might not be.
There are currently 6 billion people on earth. In their lifetimes (presuming the singularity doesn’t extend that to infinity), I expect there will be significantly less than 600 fundamental theory changes to our systems. Sure, entirely new fields like computers and genetics have occurred in the last ~80ish years, but those occurred within the framework of existing theory. Few and far between are the refutations like Phlogiston, Impetus, or Spontaneous generation. The prior probability for anyone to overturn an established branch of science is therefore already much less than 1 in 10 million.
If our prior probability is already at 1 in 10 million, additional facts like him having no training in an area and incoherent sentences push it to being even less likely.
The relevant number to use in the calculation you’re making isn’t the 6 billion people inhabiting the world, but the people that actually claim to have made such a stunning breakthrough: since that’s what instigates us to want to evaluate the claim in the first place, the fact that he made it.
In short not:
P(X made a stunning breakthrough|X is an alive human being)
but rather:
P(X made a stunning breakthrough|X claims to have made a stunning breakthrough)
As a rough guess I’d say 1⁄100 (or atleast 1/1000) may be far closer in regards to prior probabilities than 1⁄10,000,000.
“We can not solve our problems with the same level of thinking that created them”. ― Albert Einstein
If modern scientific methods could self-explain why they were wrong, they would be better scientific models. Moreover, scientific communities do not have access to perfect knowledge: any particular theory could have hundreds of supporting trials behind it if those trials weren’t popular enough to be well-known (out of millions of experiments).
OK, that’s a good point.