Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Making Vaccine

lesswrong.com

Post reply on HN
Print view View original

171–180 of 378 posts

Re: Making Vaccine

#171

Post by maxerickson »

namesbc wrote: post #168
maxerickson wrote: post #133

Earlier quoted context omitted.

Do separate branches then. Use challenge trials to accelerate deployment to under 30s (or whatever) and natural infection for other age brackets. I honestly haven't thought deeply about the ethics, but it's really concerning that it has become such a hard rule that preventing people from injuring themselves in a particular way (we don't prevent lots of other forms of injury) is somehow obviously superior to letting t…

The vaccine is going to be deployed to communities at risk first, you can't test vulnerable populations via volunteer challenge trials.

That's circular. If the challenge trial is successful, start deploying it to the groups tested using that method and then shift to the at risk groups as data justifies it.

It won't hurt the at risk groups to start up high volume manufacturing earlier (it will likely help them), and there is a benefit to everyone each time anyone is vaccinated (the reduction in transmission efficiency).

Re: Making Vaccine

#172

Post by throwaway2245 »

wtallis wrote: post #170
throwaway2245 wrote: post #163

Earlier quoted context omitted.

Does "go into the woods and gather herbs" follow the scientific method if I write my observations neatly in a blog post? The author says they will accept that their homebrew worked if they get a positive antibody test (yet, plenty of people will get a positive antibody test without any administration of homebrew or vaccine); but will not accept that it failed if there is a negative antibody test. As such, their hypot…

I think you misunderstand what "falsifiable" means. This hypothesis is absolutely falsifiable. He just doesn't have the resources to conclusively falsify this hypothesis, because he's working with a small sample size and doesn't have affordable access to testing for immunity in the mucus lining.

The hypothesis "my homebrew works at scale" is falsifiable. As you say, that really isn't under test, nor is it a hypothesis the author cares about.

The hypothesis "my homebrew has worked on me" is what the blog asks ("I'm curious whether it will work - or whether we'll be able to tell that it works."). That is not falsifiable.

Re: Making Vaccine

#173

Post by namesbc »

dcolkitt wrote: post #67
_Microft wrote: post #27

Earlier quoted context omitted.

They could not even test themselves to see whether they developed immunity against SARS-Cov2. All we know is that someone sprayed something up their nose and hoped that it actually had the effect that they wanted. This is the catch.

But the point is, even if there's a 1% chance that this plausibly works, the government should be throwing money at bio-hackers like this. Or at the very least exempting them from the regulations that stand in their way. Very likely these moonshots won't work. But they're still dirt cheap relative to the economic impact of Covid. If there's a way to disrupt the current slow approach to vaccine production, we should b…

Or, this has a 0.0001% of working, and a 20% chance of developing an autoimmune reaction that kills him.

Re: Making Vaccine

#174

Post by namesbc »

dcolkitt wrote: post #105
jiofih wrote: post #77

Earlier quoted context omitted.

If they have total immunity from FDA regulations, how are you going to be confident it actually works and is safe at scale? The regulations are there for good reasons. I haven’t heard of a single one of the available vaccines being held back by regulations, only the speed at which they could go through human trials.

> I haven’t heard of a single one of the available vaccines being held back by regulations, only the speed at which they could go through human trials. The vaccine trials were drastically slowed down by the regulations against human challenge trials.

Not true, HCT would not have speed up the deployment of the vaccine by even one day.

Re: Making Vaccine

#175

Post by wtallis »

throwaway2245 wrote: post #172
wtallis wrote: post #170

Earlier quoted context omitted.

I think you misunderstand what "falsifiable" means. This hypothesis is absolutely falsifiable. He just doesn't have the resources to conclusively falsify this hypothesis, because he's working with a small sample size and doesn't have affordable access to testing for immunity in the mucus lining.

The hypothesis "my homebrew works at scale" is falsifiable. As you say, that really isn't under test, nor is it a hypothesis the author cares about. The hypothesis "my homebrew has worked on me" is what the blog asks ("I'm curious whether it will work - or whether we'll be able to tell that it works."). That is not falsifiable.

> The hypothesis "my homebrew has worked on me" is not.

If he ends up getting COVID-19 in a few months, he'll have pretty strongly falsified that hypothesis.

You cannot dismiss a hypothesis as "not scientific" simply because it's likely that your search for evidence will be inconclusive.

Re: Making Vaccine

#176

Post by throwaway2245 »

wtallis wrote: post #175
throwaway2245 wrote: post #172

Earlier quoted context omitted.

The hypothesis "my homebrew works at scale" is falsifiable. As you say, that really isn't under test, nor is it a hypothesis the author cares about. The hypothesis "my homebrew has worked on me" is what the blog asks ("I'm curious whether it will work - or whether we'll be able to tell that it works."). That is not falsifiable.

> The hypothesis "my homebrew has worked on me" is not. If he ends up getting COVID-19 in a few months, he'll have pretty strongly falsified that hypothesis. You cannot dismiss a hypothesis as "not scientific" simply because it's likely that your search for evidence will be inconclusive.

I don't agree that is technically falsified.

I know you used "pretty strongly" as sarcasm, but falsifiability is absolute - and this is not.

(On your later point about my claim that this is not scientific, we are back to the comparison with gathering roots and berries - either both are, or both are not)

Re: Making Vaccine

#177

Post by wtallis »

throwaway2245 wrote: post #176
wtallis wrote: post #175

Earlier quoted context omitted.

> The hypothesis "my homebrew has worked on me" is not. If he ends up getting COVID-19 in a few months, he'll have pretty strongly falsified that hypothesis. You cannot dismiss a hypothesis as "not scientific" simply because it's likely that your search for evidence will be inconclusive.

I don't agree that is technically falsified. I know you used "pretty strongly" as sarcasm, but falsifiability is absolute - and this is not. (On your later point about my claim that this is not scientific, we are back to the comparison with gathering roots and berries - either both are, or both are not)

>I don't agree that is technically falsified.

> I know you used "pretty strongly" as sarcasm, but it's not falsifiability is absolute - and this is not.

I'm having a very hard time parsing what you're saying. At this point, it looks like you're trying to make an argument that "falsifiability" must mean that it is both possible and practical to reject a hypothesis with 100% certainty, and if you can't then you're not doing science. Would you care to explain more clearly why the position you're taking is less extreme and absurd than that?

(Additionally, I did not use "pretty strongly" in any sarcastic way. I used it to acknowledge the possibility of confounding factors that I did not care to enumerate, which mean that even the experimental outcome of getting COVID-19 after inoculation with the homemade vaccine would not be a 100% certain rejection of the hypothesis that the vaccine conferred some protection against COVID-19. But if you're operating in a mindset of 100% certainty being achievable and necessary for science, then I can see how you would misunderstand me.)

Re: Making Vaccine

#178

Post by namesbc »

maxerickson wrote: post #171
namesbc wrote: post #168

Earlier quoted context omitted.

The vaccine is going to be deployed to communities at risk first, you can't test vulnerable populations via volunteer challenge trials.

That's circular. If the challenge trial is successful, start deploying it to the groups tested using that method and then shift to the at risk groups as data justifies it. It won't hurt the at risk groups to start up high volume manufacturing earlier (it will likely help them), and there is a benefit to everyone each time anyone is vaccinated (the reduction in transmission efficiency).

Or, do the standard vaccine safety and efficacy testing, deploy the vaccine to the people who need it most first, and save more lives.

Optimizing deployment of the virus to bored 20 year olds may be popular among bored 20 year olds, but it isn't the best public health strategy.

Re: Making Vaccine

#179

Post by throwaway2245 »

wtallis wrote: post #177
throwaway2245 wrote: post #176

Earlier quoted context omitted.

I don't agree that is technically falsified. I know you used "pretty strongly" as sarcasm, but falsifiability is absolute - and this is not. (On your later point about my claim that this is not scientific, we are back to the comparison with gathering roots and berries - either both are, or both are not)

>I don't agree that is technically falsified. > I know you used "pretty strongly" as sarcasm, but it's not falsifiability is absolute - and this is not. I'm having a very hard time parsing what you're saying. At this point, it looks like you're trying to make an argument that "falsifiability" must mean that it is both possible and practical to reject a hypothesis with 100% certainty, and if you can't then you're not…

My point was to back up the grandparent.

This is equally science as going into the woods, picking some berries and roots, grinding them and put them up your nose, doing an antibody test, and writing your conclusions up in a blog post. There's no difference.

(But, yes, I think that writing hypotheses with 100% falsifiability, before challenging them practically, is quite a good definition of "scientific method", based on Karl Popper's work. You can compare with what Wikipedia has to say.)

Edited to respond to your edit: The word falsifiability is deliberately used to distinguish that we are talking about False, the Boolean state. The confounding factors are important - the confounding factors here mean I can't prove it False; there are no circumstances whereby the author has to accept that their hypothesis was False.

Re: Making Vaccine

#180

Post by maxerickson »

namesbc wrote: post #178
maxerickson wrote: post #171

Earlier quoted context omitted.

That's circular. If the challenge trial is successful, start deploying it to the groups tested using that method and then shift to the at risk groups as data justifies it. It won't hurt the at risk groups to start up high volume manufacturing earlier (it will likely help them), and there is a benefit to everyone each time anyone is vaccinated (the reduction in transmission efficiency).

Or, do the standard vaccine safety and efficacy testing, deploy the vaccine to the people who need it most first, and save more lives. Optimizing deployment of the virus to bored 20 year olds may be popular among bored 20 year olds, but it isn't the best public health strategy.

There is no reason to expect a challenge trial would slow down the standard branch of the trial. If it gave an earlier result, the differences would be good ones! The production would have worked through initial scaling and a bunch of people would be vaccinated.

Do you think the vaccines being distributed right now were manufactured more than a few weeks ago or something?

Post reply on HN