back to indexMichael Mina: Rapid Testing, Viruses, and the Engineering Mindset | Lex Fridman Podcast #146
link |
The following is a conversation with Michael Minna.
link |
He's a professor at Harvard doing research on infectious disease and immunology.
link |
The most defining characteristic of his approach to science and biology
link |
is that of a first principles thinker and engineer focused
link |
not just on defining the problem but finding the solution.
link |
In that spirit we talk about cheap rapid at home testing,
link |
which is a solution to COVID 19 that to me has become one of the most obvious,
link |
powerful and doable solutions that frankly should have been done months ago and still
link |
should be done now. As we talk about its accuracy is high for detecting actual contagiousness
link |
and hundreds of millions can be manufactured quickly and relatively cheaply.
link |
In general I love engineering solutions like these even if government bureaucracies often don't.
link |
It respects science and data, it respects our freedom, it respects our intelligence
link |
and basic common sense. Quick mention of his sponsor,
link |
followed by some thoughts related to the episode. Thank you to Brave,
link |
a fast browser that feels like Chrome but has more privacy preserving features,
link |
athletic greens, the all new one drink that I start every day with to cover all my nutritional
link |
bases, ExpressVPN, the VPN I've used for many years to protect my privacy on the internet,
link |
and Cash App, the app I use to send money to friends. Please check out these sponsors in
link |
the description to get a discount and to support this podcast. As a side note, let me say that
link |
I've always been solution oriented, not problem oriented. It saddens me to see that public discourse
link |
disproportionately focuses on the mistakes of those who dare to build solutions rather than
link |
applaud their attempt to do so. Teddy Roosevelt said it well in his The Man in the Arena speech
link |
over 100 years ago. I should say that both the critic and the creator are important.
link |
But in my humble estimation, there are too many now of the former and not enough of the latter.
link |
So while we spread the derisive words of the critic on social media making it viral,
link |
let's not forget that this world is built on the blood, sweat, and tears of those who dare to create.
link |
If you enjoy this thing, subscribe on YouTube, review it with five stars on our podcast,
link |
follow on Spotify, support on Patreon, or connect with me on Twitter at Lex Freedman.
link |
And now here's my conversation with Michael Mina. What is the most beautiful, mysterious,
link |
or surprising idea in the biology of humans or viruses that you ever come across in your work?
link |
Sorry for the overly philosophical question.
link |
Wow. Well, that's a great question. You know, I love the pathogenesis of viruses and
link |
one of the things that I've worked on a lot is trying to understand how viruses interact with
link |
each other. And so pre all this COVID stuff, I was really, really dedicated to understanding
link |
how viruses impact other pathogens. So how if somebody gets an infection with one thing or a
link |
vaccine, does it either benefit or harm you from other things that appear to be unrelated to most
link |
people? And so one system which is highly detrimental to humans, but what I think is just
link |
immensely fascinating is measles. And measles gets into a kid's body. The immune system picks it up
link |
and essentially grabs the virus and does exactly what it's supposed to do, which is to take this
link |
virus and bring it into the immune system so that the immune system can learn from it,
link |
can develop an immune response to it. But instead, measles plays a trick. It gets into the immune
link |
system, serves almost as a Trojan horse. And instead of getting eaten by these cells, it just
link |
takes them over and it ends up proliferating in the very cells that were supposed to kill it.
link |
And it just distributes throughout the entire body, gets into the bone marrow, kills off children's
link |
immune memories. And so it essentially, what I've found and what my research has found is that
link |
this one virus was responsible for as much as half of all of the infectious disease deaths in kids
link |
before we started vaccinating against it because it was just wiping out
link |
children's immune memories to all different pathogens, which is, I think, just astounding.
link |
It's just amazing to watch it spread throughout bodies. We've done the studies in monkeys and
link |
you can watch it just destroy and obliterate people's immune memories in the same way that
link |
some parasite might destroy somebody's brain. Is that an evolutionary coincidence or is there
link |
some kind of advantage to this kind of interactivity between pathogens? I think in that sense it's
link |
just coincidence. It's a good way for measles to essentially be able to survive long enough to
link |
replicate in the body. It just replicates in the cells that are meant to destroy it.
link |
So it's utilizing our immune cells for its own replication, but in so doing, it's destroying
link |
the memories of all the other immunological memories. But there are other viruses, so a
link |
different system is influenza and flu predisposes to severe bacterial infections. And that, I think,
link |
is another coincidence. But I also think that there are some evolutionary benefits that bacteria
link |
may hijack and piggyback on viral infections. Viruses, they just grow so much quicker than
link |
bacteria. They replicate faster. And so there's the system with viruses, with flu and bacteria,
link |
where the influenza has these proteins that cleave certain receptors. And the bacteria
link |
want to cleave those same receptors. They want to cleave the same molecules that give entrance to
link |
those receptors. So instead, the bacteria found out like, hey, you know, we could just piggyback
link |
on these viruses. They'll do it a hundred or a thousand times faster than we can. And so then
link |
they just piggyback on and they let flu cleave all these sialic acids. And then the bacteria just
link |
glom on in the wake of it. So there's all different interactions between pathogens that are just
link |
remarkable. So does this whole system of viruses that interact with each other and so damn good at
link |
getting inside our bodies, does that fascinate you or terrify you? I'm very much a scientist. And so
link |
it fascinates me much more than terrifies me. But knowing enough, I know just how well, you know,
link |
we get the wrong virus in our population, whether it's through some random mutation or
link |
whether it's this same COVID 19 virus. And these things are tricky. They're able to
link |
mutate quickly. They're able to find new hosts and rearrange in the case of influenza.
link |
So what terrifies me is just how easily this particular pandemic could have been so much
link |
worse. This could have been a virus that is much worse than it is. You know, same thing with H1N1
link |
back in 2009. That terrifies me. If a virus like that was much more detrimental, you know, that
link |
would be, it could be much more devastating. Although it's hard to say, you know, the human
link |
species were worth, well, I hesitate to say that we're good at responding to things because there
link |
are some aspects that were, this particular virus, SARS COVID 2 and COVID 19 has found a sweet spot
link |
where it's not quite serious enough on an individual level that humans just don't,
link |
we haven't seen much of a useful response by many humans. A lot of people even think it's a hoax.
link |
And so it's led us down this path of, it's not quite serious enough to get everyone to respond
link |
immediately and with the most urgency, but it's enough, it's bad enough that, you know, it's
link |
caused our economies to shut down and collapse. And so I think I know enough about virus biology to
link |
be terrified for humans that, you know, it can, it just takes one virus, just takes the wrong one
link |
to just obliterate us or not obliterate us, but really do much more damage than we've seen.
link |
It's fascinating to think that COVID 19 is a result of a virus evolving together with like
link |
Twitter, like figuring out how we can sneak past the defenses of the humans. So it's not
link |
bad enough. And then the misinformation, all that kind of stuff together is operating in such a
link |
way that the virus can spread effectively. I wonder, I mean, obviously a virus is not intelligent,
link |
but there's a, there's a rhyme and a rhythm to the way this whole evolutionary process works
link |
and creates these fascinating things that spread throughout the entire civilization.
link |
Absolutely. It's, yeah, I'm completely fascinated by this idea of social media in particular,
link |
how it replicates, how it grows, you know, I've been, how it, how it actually starts interacting
link |
with the biology of the virus, masks, who's going to get vaccinated, politics, like these seem so
link |
external to virus biology, but it's become so intertwined. And, and it's, it's interesting.
link |
And I actually think we could find out that, you know, the virus actually becomes obviously not
link |
intentionally, but, you know, we could find that choosing people choosing not to wear masks,
link |
choosing not to counter this virus in a regimented and sort of organized way effectively gives the
link |
virus more opportunity to escape. We can look at vaccines, you know, we're about to, we're about
link |
to have one of the most aggressive vaccination programs the world has ever seen. But we are
link |
unfortunately doing it right at the peak of viral transmission when millions and millions of people
link |
are still getting infected. And when we do that, that just gives this virus so many more opportunities.
link |
I mean, orders of magnitude, more opportunity to mutate around our immune system. Now, if we were
link |
to vaccinate everyone when there's not a lot of virus, then there's just not a lot of virus. And so
link |
there's not going to be as many, you know, I don't even know how many zeros are at the end of however
link |
many viral particles there are in the world right now, you know, more than quadrillions, you know,
link |
and so if you assume that at any given time, somebody might have trillions of virus in them
link |
and at any given individual. So then, you know, multiply trillions by millions. And, you know,
link |
you get a lot of viruses out there. And if you start applying pressure, ecological pressure to
link |
this virus that that, you know, when it's not abundant, kind of opportunity for a virus to
link |
sneak around immunity, especially when all the vaccines are identical, essentially.
link |
All it takes is one to mutate and then jumps. Oh, takes one takes one in the whole world, you
link |
know, and we have to, we have to not forget that this particular virus was one, it was one opportunity
link |
and it has spread across the globe and there's no reason that can't happen tomorrow. I knew,
link |
you know, it's scary. I have a million other questions in this direction, but I'd love to
link |
talk about one of the most exciting aspects of your work, which is testing or rapid testing.
link |
You wrote a great article in time on November 17th. This is like a month ago about rapid testing
link |
titled, How we can stop the spread of COVID 19 by Christmas. Let's jot down the fact that this
link |
is a month ago. So maybe your timeline would be different, but let's say in a month. So you've
link |
talked about this powerful idea for quite a while throughout the COVID 19 pandemic. How do we stop
link |
the spread of COVID 19 in a month? Well, we use tests like these. You know, so the only reason
link |
the virus continues spreading is because people spread it to each other. This isn't magic. Yes.
link |
And so there's a few ways to stop the virus from spreading to each other. And that is you either
link |
can vaccinate everyone and vaccinating everyone is a way to immunologically prevent the virus from
link |
growing inside of somebody and therefore spreading. We don't know yet actually if this vaccine, if any
link |
of these vaccines are going to prevent onward transmission. So that may or may not serve to
link |
be one opportunity. Certainly, I think it will decrease transmission. But the other idea that
link |
we have at our disposal now, we had it in May, we had it in June, July, August, September, October,
link |
November. And now it's December. We still have it. We still choose not to use it in this country and
link |
in much of the world. And that's rapid testing that is giving it's empowering people to know
link |
that they are infected and giving them the opportunity to not spread it to their loved ones
link |
and their friends and neighbors and whoever else. We could have done this. We still can.
link |
Today we could start. We have millions of these tests. These tests are simple paper strip tests.
link |
They are inside of this thing is just a little piece of paper. Now, and I can actually open it
link |
up here. There we go. So this, this is how we do it right here. We have this little paper strip test.
link |
This is enough to let you know if you're infectious with somewhere around the order of 99%
link |
sensitivity, 99% specificity, you can know if you have infectious virus in you. If we can get these
link |
out to everyone's homes, build these, make 10 million, 20 million, 30 million of them a day.
link |
You know, we make more bottles of Desani water every day. We can make these little paper strip
link |
tests. And if we do that and we get these into people's homes so that they can use them twice a
link |
week, then we can know if we're infectious, you know, is it perfect? Absolutely not. But is it
link |
near perfect? Absolutely. You know, and so if we can say, hey, the, the, the transmission of this is,
link |
you know, for every 100 people that get infected right now, they go on to infect maybe 130 additional
link |
people. And that's exponential growth. So 100 becomes 130. A couple of days later, that 130
link |
becomes another 165 people have now been infected and, you know, go over three weeks and 100 people
link |
become 500 people infected. Now, it doesn't take much to have those 100 people not infect 130,
link |
but infect 90. All we have to do is remove, say, 30, 40% of new infections from continuing their
link |
spread. And then instead of exponential growth, you have exponential decay. So this doesn't need
link |
to be perfect. We don't have to go from 100 to zero. We just have to go and have those 100 people
link |
infect 90 and those 90 people infect, you know, 82, whatever it might be. And you do that for a few
link |
weeks and boom, you have now gone instead of 100 to 500, you've gone from 100 to 20. Yes. It's not
link |
very hard. And so the way to do that is to let people know that they're infectious. I mean,
link |
I, we're a perfect example right now. I this morning, I use these tests to make sure that I
link |
wasn't infectious. Is it perfect? No, but it reduced my odds 99%. I already was at extremely low odds
link |
because I spent my life quarantining these days. Well, the interesting thing with this test,
link |
with testing in general, which is why I love what you've been espousing. And it's really
link |
confusing to me that this has not been taken on as it's one, an actual solution that's those
link |
available for a long time. There's, there doesn't seem to have been solutions proposed at a large
link |
scale and a solution that it seems like a lot of people would be able to get behind. There's some
link |
politicization or fear of other solutions that people have proposed, which is like lockdown.
link |
And there's a worry, you know, especially in the American spirit of freedom,
link |
like you can't tell me what to do. The thing about tests is it like empowers you with
link |
information essentially. So like you, it's, it gives you more information about your,
link |
like your role in this pandemic. And then you can do whatever the hell you want. Like it's all
link |
up to your ethics and so on. So like, and it's obvious that with that information, people would
link |
be able to protect their loved ones and also do, do their sort of quote unquote duty for their
link |
country, right? It's protect the rest of the country. That's exactly right. I mean, it's just,
link |
it's empowerment. But you know, this is a problem. We have not put these into action
link |
in large part because we have a medical industry that doesn't want to see them be used. We have a
link |
political and a regulatory industry that doesn't want to see them be used. That sounds crazy.
link |
Why wouldn't they want them to be used? We have a very paternalistic approach to everything in
link |
this country. You know, despite this country kind of being founded on this individualistic ideal,
link |
pull yourself up from your bootstraps, all that stuff. When it comes to public health,
link |
we have a bunch of ivory tower academics who want data. They, you know, they want to see
link |
perfection. And we have this issue of letting perfection get in the way of actually doing
link |
something at all, you know, doing something effective. And so we keep comparing these tests,
link |
for example, to the laboratory based PCR test. And sure, this isn't a PCR test, but this doesn't
link |
cost $100 and it doesn't take five days to get back, which means in every single scenario,
link |
this is the more effective test. And we have, unfortunately, a system that's not about public
link |
health. We have entirely eroded any ideals of public health in our country for the biomedical
link |
complex, you know, this medical industrial complex, which overrides everything. And that's why, you
link |
know, I'm just, can I swear on this pot? Yes. I'm just so fucking pissed that these tests don't
link |
exist. Meanwhile, and everyone says, you know, oh, we couldn't make these, you know, that we could
link |
never do it. That would be such a hard, a difficult problem. Meanwhile, the vaccine gets, we have,
link |
at the same time that we could have gotten these stupid little paper strip tests out to every
link |
household. We have developed a brand new vaccine. We've gone through phase one, phase two, phase
link |
three trials. We've scaled up its production. And now we have UPS and FedEx and all the logistics in
link |
the world getting freezers out to where they need to be. We have this immense, we see what,
link |
when it comes to sort of medicine, you know, something you're injecting into somebody,
link |
then all of a sudden people say, oh, yes, we can. But you say, oh, no, that's, that's too
link |
simple a solution, too cheap a solution. No way could we possibly do that. It's this faulty
link |
thinking in our country, which, you know, frankly is driven by big money, big, you know,
link |
the only time when we actually think that we can do something that's maybe aggressive and
link |
complicated is when there's billions and billions of billions of dollars in it, you know. And,
link |
I mean, in a difficult note, because this is part of your work from before the COVID,
link |
but it does seem that I saw statistic currently is that 40% would not be taken of Americans would
link |
not be taken a vaccine. Some number like this. So you also have to acknowledge that all the money
link |
that's been invested, like there doesn't appear to be a solution to deal with like the fear and
link |
distrust that people have. I bet, I don't know if you know this number, but for taking a strip,
link |
like a rapid test like this, I bet you people would say like the percentage of people that
link |
wouldn't take it is in the single digits probably. I completely think so. And, you know, there's a
link |
lot of people who don't want to get a test today. And that's because it gets sent to a lab,
link |
it gets reported, it has all the stuff and we're a country which teaches people from the time
link |
they're babies, you know, to keep their medical data close to them. We have HIPAA, we have all
link |
these, we have immense rules and regulations to ensure the privacy of people's medical data.
link |
And then a pandemic comes around and we just assume that all that the average person is going
link |
to wipe all of that away and say, oh, no, I'm happy giving out not just my own medical data,
link |
but also to tell the authority is everyone who I've spent my time with so that they all get a call
link |
and are pissed at me for giving up their names. You know, so people aren't getting tested and
link |
they're definitely not giving up their contacts when it comes to contact tracing. And so for
link |
so many reasons that approach is failing, not to even mention the delays in testing and things
link |
like that. And so this is a whole different approach, but it's an approach that empowers people
link |
and takes the power a bit away from the people in charge, you know, and that's what's really
link |
grating on, I think, public health officials who say, no, we need the data. So they're effectively
link |
saying, if I can't have the data, I don't want the individuals, I don't want the public to have
link |
their own data either, which is a terrible approach to a pandemic where we can't solve
link |
a public health crisis without actively engaging the public. It just doesn't work. And, you know,
link |
and that's what we're trying to do right now, which is a terrible approach.
link |
So first of all, there's, you have a really nice informed website rapidtest.org
link |
with information on this. I still can't believe this is not more popular. It's ridiculous. Okay,
link |
but our one of the FAQs you have is a rapid test too expensive. So can cost be brought down? Like,
link |
I pay, I take a weekly PCR test and I think I pay 160, 170 bucks a week.
link |
No, I mean, it's criminal. Absolutely, we can get costs. This thing right here costs less than a
link |
dollar to make with everything combined plus the swabs, you know, maybe a cost $1.50 could be sold
link |
for, frankly, it could be sold for $3 and still make a profit if they want to sell it for five,
link |
this one here. This is a slightly more complicated one, but you can see it's just got the exact
link |
same paper strip inside. And this is really, it doesn't look like much, but it's kind of the
link |
cream of the crop in terms of these rapid tests. This is the one that the US government bought
link |
and it is doing an amazing job. It has a 99.9% sensitivity and specificity. So it's really,
link |
it's really good. And so essentially the way it works is you just, you use a swab, you put the,
link |
once you, you kind of use a swab in yourself, you put the swab into these little holes here,
link |
you put some buffer on it and you close it and a line will show up if it's positive and a line
link |
won't show up, but it's negative and it takes five, 10 minutes. This whole thing, this can be made
link |
so cheap that the US government was able to buy them, buy 150 million of them from Abbott for
link |
$5 a piece. You know, so anyone who says that these are expensive, we have the proof is right here.
link |
This one at its, you know, was, Abbott did not lose money on this deal. You know, they got $750
link |
million for buying, for selling 150 million of these at five bucks a piece. All of these tests
link |
can do the same. So anyone who says that these should be, you know, unfortunately what's happening
link |
though is the FDA is only authorizing all of these tests as medical devices. So what happens when you,
link |
if I'm a medical company, if I'm, if I'm a test production company and I want to make this test
link |
and I go through and the FDA at the end of my authorization, the FDA says, okay, you know,
link |
you now have a medical device, not a public health tool, but a medical device and that
link |
affords you the ability to charge insurance companies for it. Why would I ever as a, you know,
link |
in our capitalistic economy and sort of infrastructure, why would I ever not sell this
link |
for $30 when insurance will pay for it or $100? You know, it might only cost me 50 cents to make,
link |
but, but by pushing all of these tests through a medical pathway at the FDA, they, what, what
link |
extrudes out the other side is an expensive medical device that's erroneously expensive. It doesn't
link |
need to be inflated in cost, but the companies say, well, I'd rather make fewer of them and just
link |
sell them all for $30 a piece than make tens of millions of them, which I could do and sell them
link |
at a, at a dollar marginal profit. You know, and so it's a problem with our, with our whole
link |
medical industry that we see tests only as medical devices and you know, what I would like to see
link |
is for the government in the same way that they bought 150 million of these from Abbott,
link |
they should be buying, you know, all of these tests that they should be buying 20 million a day
link |
and getting them out to people's homes. This virus has cost trillions of dollars to the American
link |
people. It's closed down restaurants and stores and, you know, obviously the main streets across
link |
America have shut, have shuttered. It's killing people. It's killing our economy. It's killing
link |
lifestyles and, and this is an obvious solution. To me, this is exciting. This is like, this is
link |
a solution. I wish, uh, like in April or something like that, uh, to launch like the larger scale
link |
manufacturing deployment of, uh, tests. It doesn't matter what tests they are. It's obviously the
link |
capitalist system would create cheaper and cheaper tests that, that would be hopefully driving down
link |
to $1. So what, what are we talking about? In America, there's, I don't know, 300 plus
link |
a million people. So that means you want to be testing regularly, right? So how many do you think
link |
it's possible to manufacture? What would be the ultimate goal to manufacture per month?
link |
Yep. So if we want to slow this virus and actually stop it from transmitting, achieve what I call
link |
herd effects like vaccine herd immunity, herd effects are when you get that R value below one
link |
through preventing onward transmission. If we want to do that with these tests, we need about
link |
20 million to 40 million of them every day, uh, which is not a lot in the United States. In the
link |
United States. So we could do it. There's other ways you can have two people in a household,
link |
swab each other, you know, swab themselves rather, and then mix, you know, put the swabs into the
link |
same tube and onto one test so you can pool. So you can get a two or three X gain in efficiency
link |
through pooling in the household. You could do that in schools or offices too, where everyone
link |
just uses a swab. You have a, there's two people, like, I mean, even if it's just standing in line
link |
at a public testing site or something, you know, you could just say, okay,
link |
these two are the last people to test or swab themselves. They go into one thing and if it
link |
comes back positive, then you just do each person and, you know, it's rapid. So you can just say
link |
to the people, one of you is positive, let's test you again. So there's ways to get the
link |
efficiency gains much better. But let's say, I think that the, that the optimal number right
link |
now that matches sort of what we can produce more or less today, if we wanted is 20 million a day.
link |
Right now, one company that I don't have their test here, but one company is already producing
link |
5 million tests themselves and shipping them overseas. It's an American company based in
link |
California called the Nova and they are giving 5 million tests to the UK every day.
link |
Not to the, you know, and this is just because there's no, the federal government hasn't
link |
authorized these tests. So without the support of the government. So yeah. So essentially,
link |
if the government just put some support behind it, then, then yeah, you can get 20 million,
link |
probably easy. Oh yeah. This, I mean, just here, I have three different companies. These,
link |
they all look similar. Well, this one's closed, but these are three different companies right here.
link |
This is a fourth Abbott, you know, this is a fifth. This is a sixth. These two are a little bit
link |
different. Do you mind if, you know, a little bit, would you take some of these or? Yeah, let's,
link |
let's do it. We can, we can, we can absolutely do them. Do you have a lot of tests in front of you?
link |
Could you maybe explain some of them? Absolutely. So there's a few different classes of tests
link |
that I just have here. And there's more tests. There's many more different tests out in the world
link |
too. These are, these are one class of tests. These are rapid antigen tests that are just
link |
the most bare bones, paper strip tests. These are, this is the type that I want to see produced in
link |
the tens of millions every day. It's so simple. You know, you don't even need the plastic cartridge.
link |
You can just, you can just make, make the paper strip and you could have a little,
link |
a little tube like this that, you know, you just dunk the paper strip into. You don't actually need
link |
the plastic, which I'd actually prefer because if we start making tens of millions of these,
link |
this becomes a lot of waste. So I'd rather not see this kind of waste be out there. And there's
link |
a few companies, Quidel is making a test called the quick view, which is just, just this. It's a,
link |
they've gotten rid of all the, all the plastic. And for people who are just listening to this,
link |
we're looking at some very small tests that fit in the palm of your hand and they're basically
link |
paper strips fit into different containers. And that's hence, hence the comment about the
link |
plastic containers. These are just injection molded, I think. And they're, you know,
link |
they can build them at high numbers, but then they have to like place them in there appropriately
link |
and all this stuff. So it is a, it is a bottleneck or some somewhat of a bottleneck in manufacturing.
link |
The actual bottleneck, which the government I think should use the defense productions act
link |
to build up is the, there's a nitrocellulose membrane, laminated membrane on this that allows
link |
the, the material, the, the, the buffer with the swab mixture to flow across it. So the way these
link |
work, they're called lateral flow tests. And you take a swab, you swab the front of your nose,
link |
you dunk that swab into some buffer, and then you put a couple of drops of that buffer onto
link |
the lateral flow. And just like paper, if you dip a piece of paper into a cup of water,
link |
though the paper will pull the water up through capillary action. This actually works very similarly
link |
at flows through, through somewhat of capillary action through this nitrocellulose membrane.
link |
And there's little antibodies on there, these little proteins that are very specific in this
link |
case for antigens or proteins of the virus. So these are antibodies similar to how, to the antibodies
link |
that our body makes from our immune system, but they're just printed on these lateral flow tests
link |
and they're printed just like a little, a line. So then you, you slice these all up into individual
link |
ones. And if there's any virus on that buffer as it flows across, the antibodies grab that virus,
link |
and it creates a little reaction with some colloids in here that cause it to turn dark.
link |
Just like a pregnancy test, one line means negative, it means the control strip worked,
link |
and two lines mean positive. It means, you know, but if you get two lines, it just means you have
link |
virus there, you're very, very likely to have virus there. And so, so they're super simple. This is,
link |
it is the exact same technology as pregnancy tests. It's the technology, this particular one
link |
from Abbott, this has been used for other infectious diseases like malaria and, and actually a number
link |
of these companies have made malaria tests that do the exact same thing. So they just coopted their,
link |
the same form factor and, and just changed the antibody. So it picks up SARS CoV2 instead of
link |
other infections. Is it also the Abbott one? Is it also a strip? Yep. Yeah. This Abbott one here is,
link |
there's the, in this case, instead of being put in a plastic sheath, it's just put in a cardboard
link |
thing and literally glued on. I mean, it's, it looks like nothing, you know, it's just, it looks
link |
like a, like, I mean, it's just the simplest thing you could, you could imagine. The exterior
link |
packaging looks very Apple like this. Nice. It does. Yeah. Yeah. Yeah. So it's nice. This is the,
link |
this is how they're packaged, you know, so and they don't have to, you know, this, these are
link |
coming in individual packages against, again, because they're really considered individual
link |
medical devices, but you could package them in, you know, bigger packets and stuff. You,
link |
you want to be careful with humidity. So they all have a little, one of those humidity removing
link |
things and oxygen removing things. So that's the, this is one class, these antigen tests.
link |
If we could just pause for a second if it's okay. And could you just briefly say what is an antigen
link |
test and what other tests there are out there, like categories of tests? Sure. Just really quick.
link |
So the testing landscape is a little bit complicated, but it's, but I'll break it down. There's really
link |
just three major classes of tests. We'll start with the first two. The first two tests are just
link |
looking for the virus or looking for antibodies against the virus. So we've heard about serology
link |
tests, or maybe some people have heard about it. Those are a different kind of tests. They're looking
link |
to see, has somebody in the past, does somebody have an immune response against the virus, which
link |
would indicate that they were infected or exposed to it. So we're not talking about the antibody
link |
tests. I'll just leave it at that. Those, they, they actually can look very similar to this or
link |
they can be done in a laboratory. They, those are usually done from blood and they're, they're
link |
looking for an immune response to the virus. So that's one. Everything I'm talking about here is
link |
looking for the virus itself, not the immune response to the virus. And so you, there's two
link |
ways to look for the virus. You can either look for the genetic code of the virus, like the RNA,
link |
just like the DNA of somebody's human cells, or you can look for the proteins themselves, the
link |
proteins themselves, the antigens of the pro of the virus. So I like to differentiate them. If you
link |
were a PCR test that looks for RNA in, let's say, let's say if we made it against humans,
link |
it would be looking for the DNA inside of our cells that would be actually looking for our genetic
link |
code. The equivalent to an antigen test is sort of a test that like actually is looking for our
link |
eyes or our nose or physical features of our body that would delineate. Okay, this is,
link |
this is Michael, for example. And so, so you're either looking for this a sequence or you're
link |
looking for a structure. The PCR tests that a lot of people have gotten now and they're done in labs
link |
usually are looking for the sequence of the virus, which is RNA. This test here by a company called
link |
Detect. This is one of Jonathan Rothberg's companies. He's the guy who helped create modern day
link |
sequencing and all kinds of other things. So this Detect device, that's the name of the company,
link |
this is actually a rapid RNA detection device. So it's almost, it's like a PCR like test,
link |
and we could even do it here. It's really, it's a beautiful test, in my opinion, works exceedingly
link |
well. It's going to be a little bit more expensive. So I think it could confirm, could be used as a
link |
confirmatory test for these. Is there a greater accuracy to it? Yes, I would say that there is a
link |
greater accuracy. There's also a downfall though of PCR and tests that look for RNA. They can
link |
sometimes detect somebody who is no longer infectious. So you have the RNA test and then you
link |
have these antigen tests. The antigen tests look for structures, but they're generally only going
link |
to turn positive if people have actively replicating virus in them. And so what happens after an
link |
infection dissipates, you have, you've just gone from having sort of a spike. So if you get infected,
link |
maybe three days later, the virus gets into exponential growth and it can replicate to
link |
trillions of viruses inside the body. Your immune system then kind of tackles it and beats it down
link |
to nothing. But what ends up in the wake of that, you just had a battle. You had this massive battle
link |
that just took place inside your upper respiratory tract. And because of that, you've had trillions
link |
and trillions of viruses go to zero, essentially. But the RNA is still there. It's just these remnants
link |
in the same way that if you go to a crime scene and blood was sort of spread all over the crime
link |
scene, you're going to find a lot of DNA. There's tons of DNA. There's no people anymore,
link |
but there's a lot of DNA there. Same thing happens here. And so what's happening with PCR testing
link |
is when people go and use these exceedingly high sensitivity PCR tests, people will stay positive
link |
for weeks or months after their infection has subsided, which has caused a lot of problems in
link |
my opinion. It's problems that the CDC and the FDA and doctors don't want to deal with. But I've
link |
tried to publish on it. I've tried to suggest that this is an issue both in New York Times
link |
and others. And now it's unfortunately kind of taken on a life of its own of conspiracy theorists
link |
thinking that they call it a case demig. They say, oh, PCR is detecting people who are false
link |
positive. They're not false positives. They're late positives, no longer transmissible.
link |
I think the way you, like what I saw in rapidest.org, I really like the distinction between
link |
diagnostic sensitivity and contagiousness sensitivity. That's so, that website is so obvious
link |
that it's painful because it's like, yeah, that's what we should be talking about is
link |
how accurately is a test able to detect your contagiousness? And you have different plots
link |
that show that actually there's, you know, that antigen tests, the tests we're looking at today
link |
like rapid tests are actually really good at detecting contagiousness. Absolutely. It all
link |
mixes back with this whole idea that of the medical industrial complex, you know, in this country
link |
and in most countries, we have almost entirely defunded and devalued public health period. You
link |
know, we just have. And what that means is that we don't even, we don't have a language for it.
link |
We don't have a lexicon for it. We don't have a regulatory landscape for it. And so the only
link |
window we have to look at a test today is as a medical diagnostic test. And that becomes very
link |
problematic when we're trying to tackle a public health threat and a public health emergency by
link |
definition. This is a public health emergency that we're in. And yet we keep evaluating tests
link |
as though the diagnostic benchmark is the gold standard where if I'm a physician, I am a physician,
link |
so I'll put on that physician hat for a moment. And if I have a patient who comes to me and wants to
link |
know if their symptoms are a result of them having COVID, then I want every shred of evidence that
link |
I can get to see, does this person currently or did they recently have this infection inside of them?
link |
And so in that sense, the PCR test is the perfect test. It's really sensitive. It will find the RNA
link |
if it's there at all so that I could say, you know, yeah, you have a low amount of RNA left. You
link |
might have been, you said your symptoms started two weeks ago. You probably were infectious two
link |
weeks ago and you have lingering symptoms from it. But that's a medical diagnosis. It's kind of
link |
like a detective recreating a crime scene. They want to go back there and recreate the pieces
link |
so that they can assign blame or whatever it might be. But that's not public health. And public
link |
health, we need to only look forward. We don't want to go back and say, well, was this person,
link |
are there symptoms because they had an infection two weeks ago? In public health, we just want to
link |
stop the virus from spreading to the next person. And so that's where we don't care if somebody was
link |
infected two weeks ago. We only care about finding the people who are infectious today. And unfortunately,
link |
our regulatory landscape fails to apply that knowledge to evaluate these tests as public health
link |
tools. They're only evaluating the tests as medical tools. And therefore, we get all kinds of
link |
complaints that say this test, which detects 99 plus, 99.8% of current infectious people.
link |
By the FDA's rubric, they'll say, no, no, it's only 50% sensitive. And that's because when you go out
link |
into the world and you just compare this against PCR positivity, most people who are PCR positive
link |
in the world right now at any given time, are post infectious. They're no longer infectious
link |
because you might only be infectious for five days, but then you'll remain PCR positive for
link |
three or four or five weeks. And so when you go and just evaluate these tests and you say, okay,
link |
this person's PCR positive, does the rapid antigen test detect that? More often than not,
link |
it's no. But that's because those people don't need isolation. They're post infectious. And this is
link |
a, it's become much more of a problem than I think even the FDA themselves is recognizing
link |
because they are unwilling at this point to look at this as a public health problem requiring
link |
public health tools. We'll definitely talk about this a little bit more because the concern I have
link |
is that like a bigger pandemic comes along, what are the lessons we draw from this and how we move
link |
forward? Let's talk about that in a bit. But sort of, can we discuss further the lay of the land
link |
here of the different tests before us? Absolutely. So I talked about PCR tests and those are done in
link |
the lab or they're done essentially with a rapid test like this, the detect. And we can even try
link |
this in a moment. It goes into a little heater. So you might have one of these in a household or
link |
one of these in a nursing home or something like that, or in an airport,
link |
or you could have one that has 100 different outlets. This is just to heat the tube up.
link |
These are the rapid tests. They are super simple. No frills. You just swab your nose and you put the
link |
swab into a buffer and you put the buffer on the test. So we can use these right now if you want.
link |
Yeah. We can try it out. And all the tests we're talking about, they're usually swabbing the nose.
link |
Like that's the... That's still the main, yeah. There are some saliva tests coming about and these
link |
can all work potentially with saliva. They just have to be recalibrated. But these swabs are really
link |
not bad. This isn't the deep swab that goes like way back into your nose or anything. This is just
link |
a swab that you do yourself like right in the front of your nose. So if you want to do it...
link |
Yeah. Do you mind if I...
link |
Sure. Yeah. Yeah. Why don't we start with this one? Because this is the Abbots Bynex Now test
link |
and it's really... It's pretty simple. This is the swab from the Abbott test.
link |
That's correct. That's the swab from the Abbott test. So what I'm going to do to start
link |
is I'm going to take this buffer here, which is... This is just the buffer that goes onto this test.
link |
So this is a brand new one. I just opened this test out. I'm going to just take six drops of this
link |
buffer and put it right onto this test here. Two, three, four, five, six.
link |
Okay. And now you're going to take that swab, open it up.
link |
Yep. And now just wipe it around inside the... Into the front of your nose.
link |
Do a few circles on each nostril.
link |
Yeah. That looks good.
link |
This always makes me want to sneeze.
link |
Yeah. Okay. Now I'm going to have you do it yourself.
link |
I'm getting emotional.
link |
Hold it parallel to the test. So put the test down on the table. Yep. And then go into that
link |
bottom hole. Yep. And push forward so that you can start to see it in the other hole.
link |
There you go. Now turn. If it's... Once it hits up against the top, just turn it three times.
link |
One, two, three, and sort of... Yep. And now you just close. So pull off that adhesive sticker there.
link |
And now you just close the whole thing.
link |
And... And that's it. That's it. Now what we will see is we will see a line form.
link |
What's happening now is the buffer that you put in there is now
link |
moving up onto the paper strip test. And it has the material from the swab in there.
link |
And so what we'll see is a line will form. And that's going to be the control line.
link |
And then we'll also see the... Ideally, we'll see no line for the actual test line.
link |
And that's because you should be negative. So one line will be positive and two lines will be negative.
link |
That's very cool. There's this purple thing creeping up onto the control line.
link |
That's perfect. That's what you want to be seeing. So you want to see that...
link |
So right now, you essentially want to see that that blue line turns pink or purpley color.
link |
There's a blue line that's already there printed. It should turn sort of a purple pink color.
link |
And ideally, there will be no additional line for the sample.
link |
And if there is, that's the 99. whatever percent accuracy on... That means I have... I'm contagious.
link |
That would mean that you're likely contagious or you likely have infectious virus in you.
link |
What we can do, because one of the things that my plan calls for,
link |
is because sometimes these tests can get false positive results. It's rare.
link |
Maybe 1% or in the case of this by next now, this Abbott test, 0.1%. So 1 in 1,000, 1 in 500,
link |
something like that can be falsely positive. What I recommend is that when somebody is
link |
positive on one of these, you turn around and you immediately test on a different test.
link |
You could either do it on the same, but for a good measure, you want to use a separate test
link |
that is somewhat orthogonal, meaning that it shouldn't turn falsely positive for the same reason.
link |
This particular test here, this detect test, because it is looking for the RNA and not the
link |
antigen, this is an amazingly accurate test. And it's sort of a perfect gold standard or
link |
confirmatory test for any of these antigen tests. So one of the recommendations that I've had,
link |
especially if people start using antigen tests before you get onto a plane or as what I call
link |
entrance screening, if somebody's positive, you don't immediately tell them you're positive,
link |
go isolate for 10 days. You tell them, let's confirm on one of these on a detect test.
link |
That is because it's completely orthogonal, it's looking for the RNA instead of the antigen. There's
link |
no reason, no biological reason that both of these should be falsely positive. So if one's
link |
falsely positive and the other one is negative, especially because this one's more sensitive,
link |
then I would trust this as a confirmatory test. If this one's negative, then the antigen test
link |
would be considered falsely positive. It does look like there's only a single line. This is very
link |
exciting news. That's right. Yep. It says wait 15 minutes to see both lines. But in general,
link |
if somebody's really going to be positive, that line starts showing up within a minute or two.
link |
So you want to keep the whole, we'll keep watching it for the whole 15 minutes as it's
link |
sitting there. But I would say you're knowing that you've had PCR tests recently and all that.
link |
The odds are pretty good. The odds are very good. The packaging, very iPhone like,
link |
I'm digging the sexy packaging. I'm a sucker for good packaging. Okay.
link |
So then there's this test here, which is, this is another, it's funny. Let me open this up and
link |
show you. This is a really nice test. It's another antigen test. Works the exact same way as this,
link |
essentially. But what you can see is it's got lights in it and a power button and stuff. This
link |
is called an allume test, which is fine. And it's a really nice test, to be honest. But it has to
link |
pair with an iPhone. And so it's good as a, I think that this is going to become, there's a
link |
lot of use for this from a medical perspective, where you want good reporting. This can, because
link |
it pairs with an iPhone, it can immediately send the report to a Department of Health,
link |
whereas these paper strip tests that they're just paper, they don't report anything unless you
link |
want to report it. Okay. So I'm going to just pick it up and pick it apart. And so you can see,
link |
is there's like fluorescent readers and little lasers and LEDs and stuff in there.
link |
You can actually see the lights going off. And there's a paper strip test right inside there,
link |
but you can see that there's like a whole circuit board and all this stuff. Right. And so
link |
this is the kind of thing that, you know, the FDA is looking for, for like home use and things like
link |
that, because it's kind of foolproof. Like you can't go wrong with it. It pairs with an iPhone,
link |
so you need Bluetooth. So it's going to be more limited. It's a great test. Don't get me wrong.
link |
It's as good as any of these. But, you know, when you compare this thing with a battery and a
link |
circuit board and all this stuff, it's got its purpose, but, you know, it's not a public health
link |
tool. I don't want to see this made in the tens of millions a day and thrown away.
link |
But FDA likes that kind of stuff. FDA loves this stuff, you know, because they can't get
link |
it out of their mind that this is a public health crisis. You know, we need, we need,
link |
I mean, just look at the difference here. Something flashing lights is essential.
link |
Got batteries. It's got a Bluetooth thing. It's a great test, but, you know, it's,
link |
to be honest, it's not any better than this one. And so, you know, I want this one.
link |
It's nice and all. The form factor is nice, but, and it's really nice that it goes to Bluetooth.
link |
But it goes against the principle of just 20 million a day. The easy solution. Everybody has it.
link |
You can manufacture and probably, you could have probably scaled this up in a couple of weeks.
link |
Oh, absolutely. These companies, I mean, the rest of the world has these. They can be scaled up.
link |
They already exist. You know, SD biosensors, one company is making tens of millions a day,
link |
not coming to the United States, but going all over Europe, going all over Southeast Asia and
link |
East Asia. So they exist. The US is just, you know, we can't get out of our own way.
link |
I wonder what, why somebody, I don't know if you were paying attention, but somebody like an Elon
link |
Musk type character. So he was really into doing some like obvious engineering solution, like this
link |
at home rapid test seems like a very Elon Musk thing to do.
link |
Well, I don't know if you saw, but I had a little Twitter conversation with Elon Musk.
link |
Does he not like what was it? Do you know what his thoughts are on rapid testing?
link |
Well, he was using a slightly different one, one of these, but that requires an instrument called
link |
the BD Veritor. And he got a false positive or no, I shouldn't say he didn't necessarily get a
link |
false positive. He got discrepant results. He did this test four times. He got two positives, two
link |
negatives. But then he got a PCR test and it was a very low positive result. So I think
link |
what happened is he just tested himself at the tail end of it. This was actually right before
link |
he was about to send those. It was the day of essentially that he was sending the astronauts
link |
up to the space station the other day. So he was using these rapid tests because he wanted to
link |
make sure that he was good to go in and he got discrepant results. Ultimately, they were correct,
link |
but you know, two were negative, two were positive. But what really happened, once he got his,
link |
he shared his PCR results and they were very low positive. So really what was happening
link |
is my guess is he found himself right at the edge of his positivity, of his infectiousness.
link |
And so, you know, the test worked out, it was supposed to work. It probably had he used it
link |
two days earlier. It would have been screaming positive. You know, he wouldn't have gotten
link |
discrepant results. But he found himself right at the edge by the time he used the test. So the
link |
PCR would always pick it up because it's still, because it will still stay positive then for
link |
weeks, potentially. But the rapid antigen test was starting to falter, not in a bad way, but just
link |
he probably was really no longer particularly infectious. And so it was kind of when it gets
link |
to be a very low viral load, it becomes stochastic.
link |
It's fascinating this duality. So one, you can think from an individual perspective,
link |
it's unclear when you take four and a half are positive, half a negative, like what are you
link |
supposed to do? But from a societal perspective, it seems like if just one of them is positive,
link |
just stay home for a couple days, for a while. So if one year CEO of a company you're launching
link |
astronauts to space, you may not want to rely absolutely on the antigen test as a thing by
link |
which you steer your decisions of like 10,000 plus people companies. But us individuals just
link |
living in the world, if it comes up positive, then you make decisions based on that. And then
link |
that scales really nicely to an entire society of hundreds of millions of people. And that's how
link |
you get that virus to a stop spreading. That's exactly right. You don't have to catch every
link |
single one. And the nice thing is that these will, these will catch the people who are most
link |
infectious. So with Elon Musk, generally that test, we don't have the counterfactual. We don't
link |
have his results from three days earlier when he was probably most infectious. But my guess is the
link |
fact that it was catching two out of the four, even when he was down at a CT value of really,
link |
really very, very low viral load on the PCR test suggests that it was doing its job. And you just
link |
want to, and the nice thing is because these can be produced at such scale, getting one positive
link |
doesn't immediately have to mean 10 days of isolation. That's the CDC's more conservative
link |
stance to say, if you're positive on any test, stay home for 10 days and isolate. But here,
link |
people would just have more tests. So the recommendation should be test daily. If you
link |
turn positive test daily, until you've been negative for 24, 48 hours, and then go back to
link |
work. And the nice thing there is, right now, people just aren't testing because they don't want
link |
to take 10 days off. They're not getting paid for it. So they can't take 10 days off.
link |
Do you know what Elon thinks about this idea of rapid testing for everybody? So I understood,
link |
I need to look at that whole Twitter thread. So I understand his perhaps criticism of,
link |
he had like a conspiratorial tone for my vague look at it of like, what's going on here with
link |
these tests? But what does he actually think about this very practical to me engineering
link |
solution of just deploying rapid tests to everybody? It seems like that's a way to open
link |
up the economy in April. Well, to be honest, I've been trying to get in touch with him again.
link |
I think, take somebody like Elon Musk with the engineering prowess within his ranks to
link |
easily, easily build these at the tens of millions a day. He could build the machines from scratch.
link |
A lot of the companies, they buy the machines from South Korea or Taiwan, I believe. We don't
link |
have to. We can build these machines. They're simple to build. Put somebody like Elon Musk on it.
link |
Take some of his best engineers and say, look, the US needs a solution in two weeks. Build
link |
these machines. Figure it out. He'll do it. He could do it. This is a guy who is literally,
link |
he started multiple entirely new industries. He has the capital to do it without the US government
link |
if he wanted to. The return on investment for him would be huge. But frankly, the return on
link |
investment in the country would be hundreds of billions of dollars because it means we could
link |
get society open. I know that his first experience with these rapid tests was confusing, which is
link |
how I ended up having this Twitter conversation with him very briefly. But I think that if he
link |
understood a little bit more, and I think he does, I really love to talk to him about it,
link |
because I think he could totally change the course of this pandemic in the United States
link |
singlehandedly. He loves grand things. I think out of all the solutions I've seen,
link |
this is the obvious engineering solution to at least a pandemic of this scale.
link |
I love that you say the engineering solution. So this is something I've been really trying to,
link |
I'm an engineer. My previous history was all engineering, and that's really how I think.
link |
I then went into medicine and PhD world. But I think that the world, one of the major
link |
catastrophes or one of the major problems is that we have physicians making the decisions about
link |
public health and a pandemic when really we need engineers. This is an engineering problem.
link |
What I've been trying to do, I actually really want to start a whole new field called public
link |
health engineering. Eventually, I want to try to bring it to MIT and get MIT to want to start a
link |
new department or something. That's a doubly awesome idea. I love this. I love every aspect.
link |
I love everything you're talking about. A lot of people believe because vaccines started being
link |
deployed currently that we are no longer in need of a solution. We're no longer in need of slowing
link |
the spread of the virus. To me, as I understand, it seems like this is the most important time
link |
to have something like a rapid testing solution. Can you break that apart? What's the role of rapid
link |
testing in the next three, four months maybe? Even more. The vaccine rollout isn't going to be as
link |
peachy as everyone is hoping. I hate to be the Debbie Downer here, but there's a lot of unknowns
link |
with this vaccine. You've already mentioned one, which is there's a lot of people who just don't
link |
want to get the vaccine. I hope that that might change as things move forward and people see
link |
their neighbors getting it and their family getting it, and it's safe and all. We don't know how
link |
effective the vaccine is going to be after two or three months. We've only measured it in the first
link |
two or three months, which is a massive problem, which we can go into biologically because there's
link |
reasons, very good reasons to believe that the efficacy could fall way down after two or three
link |
months. We don't know if it's going to stop transmission, and if it doesn't stop transmission,
link |
then herd immunity is much, much more difficult to get because that's all based on transmission
link |
blockade. And frankly, we don't know how easily we're going to be able to roll it out. Some of
link |
the vaccines need really significant cold chains, have very short half lives outside of that cold
link |
chain. We need to organize massive numbers of people to be able to distribute these. Most
link |
hospitals today are saying that they're not equipped to hire the right people to be even
link |
administering enough of these vaccines. And then a lot of the hospitals are frustrated because
link |
they're getting much lower, smaller allocations than they were expecting. So I think right now,
link |
like you say, right now is the best time, besides three or four or five or six months ago,
link |
right now is the best time to get these rapid tests out. The country has the capacity to build
link |
them. We're shipping them overseas right now. We just need to flip a switch, get the FDA to
link |
recognize that there's more important things than diagnostic medicine, which is the effectiveness of
link |
the public health program when we're dealing with a pandemic. They need to authorize these as public
link |
health tools or frankly, the president could. There's a lot of other ways to get these tests to
link |
not have to go through the normal FDA authorization program, but maybe have the NIH and the CDC
link |
give a stamp of approval. And if we could, we could get these out tomorrow. And that's where
link |
that article came from, how we can stop the spread of this virus by Christmas. We could. Now it's
link |
getting late. And so we have to keep updating that timeframe, maybe putting Christmas in the title
link |
wasn't, I should have said how we can stop the spread of this virus in a month. It would be a
link |
little bit more timeless, but we could do it. We really could do it. And that's the most frustrating
link |
part here is that we're just choosing not to as a country. We're choosing to bankrupt our society
link |
because some people at the FDA and other places just can't seem to get their head around the fact
link |
that this is a public health problem, not a bunch of medical problems. Is there a way to change that
link |
policy wise? So this is, this is a much bigger thing that you're speaking to, which I, I love in
link |
terms of the MIT engineering approach to public health. Is there a way to push this? Is this,
link |
is this a political thing like where some Andrew Yang type characters need to like start screaming
link |
about it? Is it more of an Elon Musk thing where people just need to build it and then on Twitter
link |
start talking crap to politicians who are not doing it? What are the ideas here?
link |
I think it's a little both. I think it's political on the one hand, and I've certainly been talking
link |
to Congress a lot, talking to senators. Are they receptive? Oh, yeah. I mean, that's the crazy
link |
thing. Everyone but the FDA is receptive. I mean, it's, it's astounding. I mean, I advise, you know,
link |
informally, I advise the president and the president elects teams. I talk to Congress,
link |
I talk to senators, governors, you know, and then all the way down to mayors of towns and things.
link |
And I held, I mean, months ago, I held a roundtable discussion with Mayor Garcetti,
link |
who's the mayor of LA. And I brought all the, all the companies who make these things. This was in
link |
like July or August or something. I brought all the companies to the table and said, okay, how can
link |
we get these out? And unfortunately, it went nowhere because the FDA won't authorize them as
link |
public health tools. The nice thing is that this is one of the nice and frustrating things. This
link |
is one of the few bipartisan things that I know of. And like you said, it's, it's a real solution.
link |
Lockdowns aren't a solution. They're, they're an emergency bandaid to a catastrophe that's
link |
currently happening. They're not a solution. And they're definitely not a public health solution
link |
if we're taking a more holistic view of public health, which includes people's well being,
link |
includes their psychological well being, their financial well being, you know,
link |
just stopping a virus if it means that all those other things get thrown under the bus
link |
is not a public health solution. It's a, it's a, it's a myopic or very tunnel vision
link |
approach to a viral virus that's spreading. This is a simple solution with essentially
link |
no downfall. You know, there is no, nothing bad about this. It's just giving people
link |
a result. And it's bipartisan, you know, the most conservative and the most liberal people.
link |
Everyone just wants to know their status. You know, nobody wants to have to wait in line for
link |
four hours to find out their status on Monday, a week later on Saturday. You know, it just doesn't
link |
make any sense. It's a useless test at that point. And everyone recognizes that.
link |
So why, why do you think like the mayor of LA, why do you think politicians are going for these,
link |
from my perspective, like kind of half ass lockdowns, which is not, so I have seen good
link |
evidence that like a complete lockdown can work, but that's in theory, it's just like
link |
communism in theory can work. Like theoretically speaking, but it just doesn't, at least in
link |
this country, we don't, I think it's just impossible to have complete lockdown. And still,
link |
politicians are going for these kind of lockdowns that everybody hates,
link |
that's really, really hurting small businesses. Like, why are they going for that?
link |
Yeah, all businesses, but like basically not just hurting, they're destroying small businesses,
link |
right? Which is going to have potentially, I mean, yeah, I've been reading as I don't
link |
shout out about the rise and fall of the third Reich. And, you know, there's economic effects
link |
that take a decade to, you know, there's going to be long lasting effects that may maybe
link |
destructive to the very fabric of this nation. So why are they doing it? And why they're not
link |
using the solution? Is there an intuition? I mean, you've said that FDA has a stranglehold,
link |
I guess, on this whole public health problem. Is that all it is?
link |
That's honestly, it's pretty much all it is. The company, so the somebody like Mayor Garcetti or
link |
Governor Baker, Cuomo Newsom, any of these, DeWine, I've talked to a lot of governors in
link |
this country at this point. And of course, the federal government, including the president's
link |
own teams, you know, and the heads of the NIH, the heads of the CDC about this.
link |
The problem is the tests don't exist in this country at the level that we need them to right now
link |
to make that kind of policy, to make that kind of program. They could, but they don't. And so
link |
what that means is that when Mayor Garcetti says, okay, what are my actual options today,
link |
despite these sounding like a great idea, he looks around and he says, well, they're not
link |
authorized. You know, they don't exist right now for at home use. And from his perspective,
link |
he's not about to pick that fight with the FDA. And it turns out nobody is.
link |
Why are people afraid of it? It seems like an easy struggle to fight. It's like, uh...
link |
Well, it's not a... So they don't see it as a fight. They think that the FDA is the end all
link |
be all. Everyone thinks the FDA is the end all be all. And so they just differ. Everyone is
link |
deferential, including the heads of all the other government agencies, because that is their role.
link |
But what everyone is failing to see is that the FDA doesn't even have a mandate or a remit
link |
to evaluate these tests as public health tools. So they're just falling in this weird gray zone
link |
where the FDA is saying, look, we evaluate medical products. That's the only thing that
link |
I meant, like Tim Stenzel, head of Invitra Diagnostics at the FDA. He's doing what his job is,
link |
which is to evaluate medical tools. Unfortunately, this is where I think the CDC has really
link |
blended. They haven't made the right distinction to say, look, okay, the FDA is evaluating these
link |
for doctors to use and all that. But we're the CDC and we're the public health agency of this
link |
country. And we recognize that these tools require a different authorization pathway and a different
link |
use, not prescription. There's a difference in medical devices and public health. I guess FDA is
link |
not designed for this public health, especially in emergency situations. And they actually
link |
explicitly say that. I mean, when I go and talk to Tim, he's a very reasonable guy.
link |
But when I talk to him, he says, look, we don't, we just do not evaluate a public health tool. If
link |
you're telling me this is a public health tool, great, go and use it. And so I say, okay, great,
link |
we'll go and use it. And then the comment is, but does it give a result back to somebody? I say,
link |
well, yes, of course it gives a result back to somebody. It's being done in their home. So
link |
well, then it's defined as a medical tool. Can't use it. So it's stuck in this gray zone where
link |
unfortunately, there's this weird definition that any tool, any test that gives a result back to
link |
an individual is defined by CMS, Centers for Medicaid Services as a medical device requiring
link |
medical authorization. But then you go and ask, it gets crazier because then you go and ask
link |
Seema Verma, the head of CMS, okay, can these be authorized as public health tools and not
link |
fall under your definition of a medical device? So then the FDA doesn't have to be the ones authorizing
link |
it as a public health tool. And Seema Verma says, oh, well, we don't have any jurisdiction over
link |
point of care and sort of rapid devices like this. We only have jurisdiction over lab devices.
link |
So it's like nobody has ownership over it, which means that they just keep, they stay in this purgatory
link |
of not being approved. And so this is where I think, frankly, it needs a president. It needs a
link |
presidential order to just unlock them to say, this is more important than having a prescription.
link |
And in fact, I mean, really what's happening now because there is this sense that tests are public
link |
health tools, even if they're not being defined as such, the FDA now is pretty much, not only are
link |
they not authorizing these as public health tools, what they're doing by authorizing what are
link |
effectively public health tools as medical devices, they're just diluting down the practice of medicine.
link |
Right. I mean, his answer right now, unfortunately, is, well, I don't know why you want these to be
link |
sort of available to everyone without a prescription. We've already said that a doctor can write
link |
a whole prescription for a whole college campus. It's like, well, if you're going in that direction,
link |
then that's no longer medicine. Having a doctor write a prescription for a college campus for
link |
everyone on the campus to have repeat testing, now we're just in the territory of eroding medicine
link |
and eroding all of the legal rules and reasons that we have prescriptions in the first place.
link |
So it's just everything about it is just destructive instead of just making a simple solution,
link |
which is, these are okay as public health tools as long as they meet X and Y metrics go and CDC can
link |
put their stamp of approval on them. What do you think? Sorry if I'm stuck on this. You mentioned
link |
of MIT and public health engineering, right? I mean, it has a sense of I talked to competition
link |
biology folks. It's always exciting to see computer scientists start entering the space of biology.
link |
And there's actually a lot of exciting things that happen because of that, trying to understand
link |
the fundamentals of biology. So from the engineering approach to public health,
link |
what kind of problems do you think can be tackled? What kind of disciplines are involved?
link |
Like, do you have ideas in this space? Oh, yeah. I mean, I can speak to one of the major
link |
activities that I want to do. So what I normally do in my research lab is develop
link |
technologies that can take a drop of somebody's blood or some saliva and profile for hundreds of
link |
thousands of different antibodies against every single pathogen that somebody could be possibly
link |
exposed to. So this is all new technology that we've been developing more from a bioengineering
link |
perspective. But then I use a lot of the mathematics tools to interpret that. But what I really want
link |
to do, for example, to kind of kick off this new field of what I consider public health
link |
engineering is to create, maybe it's a little ambitious, but create a weather system for viruses.
link |
I want us to be able to open up our iPhones, plug in our zip code, and get a better sense,
link |
get a probability of why my kid has a runny nose today. Is it COVID? Is it a rhinovirus?
link |
An adenovirus or is it flu? And we can do that. We can start building the rules of virus spread
link |
across the globe, both for pandemic preparedness, but also for just everyday use. In the same way
link |
that people used to think that predicting the weather was going to be impossible. Of course,
link |
we know that's not impossible now. Is it always perfect? No. But does it offer, does it completely
link |
change the way that we go about our days? Absolutely. I envision, for example, right now,
link |
we open up our iPhone, we plug in a zip code, and if it tells us it's going to rain today,
link |
we bring an umbrella. So in the future, it tells us, hey, there's a lot of SARS code too in your
link |
community. Instead of grabbing your umbrella, you grab your mask. We don't have to have masks all the
link |
time. But if we know the rules of the game that these viruses play by, we can start preparing
link |
for those. And every year, we go into every flu season blindfolded with our hands tied behind
link |
our back, just saying, I hope this isn't a bad flu season this year. We're in the 21st century.
link |
We have the tools at our disposal now to not have that attitude. This isn't like
link |
1920s. We can just say, hey, this is going to be a bad flu season this year. Let's act accordingly
link |
and with a targeted approach. For example, we don't just use our umbrellas all day long every
link |
single day in case it might rain. We don't board up our homes every single day in case it's a
link |
hurricane. We wait, and if we know that there's one coming, then we act for a small period of
link |
time accordingly. And then we go back and we've prepared ourselves in these little bursts to
link |
not have it ruin our days. I can't tell you how exciting that vision of the future is. I think
link |
that's incredible. And it seems like it should be within our reach. Just these weather maps of
link |
viruses floating about the earth. And it seems obvious. It's one of those things where right now
link |
it seems like maybe impossible. And then looking back like 20 years from now, we'll wonder why
link |
the hell this hasn't been done way earlier. Though one difference between weather, I don't know if
link |
you have interesting ideas in this space. The difference between weather and viruses is it
link |
includes the collection of the data, includes the human body potentially. And that means that
link |
there is some, as with the contact tracing question, there's some concern about privacy.
link |
There seems to be this dance that's really complicated with Facebook getting a lot of
link |
flack for basically misusing people's data or just whether it's perception or reality,
link |
there's certainly a lot of reality to it too, whether they're not good stewards of our private
link |
data. So there's this weird place where it's obvious that if we collect a lot of data about
link |
human beings and maintain privacy and maintain all basic respect for that data, just like
link |
honestly common sense respect for the data, then we can do a lot of amazing things for the world,
link |
like a weather map for viruses. Is there a way forward to gain trust of people
link |
or to do this well? Do you have ideas here? How big is this problem?
link |
I think it's the central problem. There's a couple central problems that need to be solved.
link |
One, how do you get all the samples? That's not actually too difficult. I actually have a pilot
link |
project going right now with getting samples from across all the United States. Tens of thousands
link |
of samples every week are flowing into my lab and we process them. So it's taking like one of the
link |
basically this biology here in chemistry and converting that into numbers. That's exactly
link |
right. So what we're doing, for example, there's a lot of people who go to the hospital every day,
link |
a lot of people who donate blood, people who donate plasma. So one of the projects that I have,
link |
I'll get to the privacy question in a moment, but this, so what I want to do is that the name
link |
that I've given this is a global immunological observatory. There's no reason not to have that.
link |
Good name. I've said, instead of saying, well, how do we possibly get enough people on board to
link |
send in samples all the time? Well, just go to the source. So there's a company in Massachusetts
link |
that makes 80% of all the instruments that are used globally to collect plasma from plasma donors.
link |
So I went to this company, Hemenetics, and said, is there a way you have 80% of the global market
link |
on plasma donations? Can we start getting plasma samples from healthy people that use your machines?
link |
So that hooked me up with this company called Octafarma. And Octafarma has a huge reach
link |
and offices all over the country where they're just collecting people's plasma. They actually pay
link |
people for their plasma, and then that gets distributed to hospitals and all this stuff
link |
is anonymous plasma. So I've just been collecting anonymous samples. And we're processing them,
link |
in this case, for COVID antibodies to watch from January up through December, we're able to
link |
watch how the virus entered into the United States and how it's transmitting every day
link |
across the US. So we're getting those results organized now, and we're going to start
link |
putting them publicly online soon to start making at least a very rough map of COVID.
link |
But that's the type of thinking that I have in terms of like, how do you actually capture huge
link |
numbers of specimens? You can't ask everyone to participate on sort of a, I mean, you maybe
link |
could if you have the right tools, and you can offer individuals something in return like 23 in
link |
me does. That's a great way to get people to give specimens and they get results back. So
link |
with these technologies that I've been building, along with some collaborators at Harvard, we can
link |
come up with tools that people might actually want. So I can offer you your immunological history. I
link |
can say, give me a drop of your blood on a filter paper, mail it in, and I will be able to tell you
link |
every infectious disease you've ever encountered, and maybe even when you encountered it roughly.
link |
I could tell you, do you have COVID antibodies right now? Do you have Lyme disease antibodies
link |
right now? Flu, triple E, and all these different viruses, also peanut allergies, milk allergies,
link |
anything. If your immune system makes a response to it, we can detect that response.
link |
So all of a sudden, we have this very valuable technology that on the one hand,
link |
gives people maybe information they might want to know about themselves. But on the other hand,
link |
becomes this amazingly rich source of big data to enter into this global immunological observatory
link |
sort of mathematical framework to start building these maps, these epidemiological tools. But
link |
you ask about privacy. And absolutely, that's essential to keep in mind, first and foremost.
link |
So privacy can be, you can keep these samples 100% anonymous. They're just, when I get them,
link |
they show up with nothing. They're literally just tubes. I know a date that they were collected and
link |
a zip code that they're collected from or even just sort of a county level ID. With an IRB and
link |
with ethical approval and with the people's consent, we can maybe collect more data,
link |
but that would require consent. But then there's this other approach, which I'm really excited
link |
about, which is certainly going to gain some scrutiny, I think. But we'll have to figure out
link |
where it comes into play. But I've been recognizing that we can take somebody's
link |
immunological profile, and we can make a biological fingerprint out of it. And it's
link |
actually stable enough so that I could take your blood. Let's say I don't know who you are,
link |
but you sent me a drop of blood a year ago, and then you sent me a drop of blood today.
link |
I don't know that those two blood spots are coming from the same person. They're just showing up
link |
in my lab. But I can run our technology over, and it just gives me your immunological history.
link |
But your immunological history is so unique to you, and the way that your body responds to these
link |
pathogens is so unique to you that I can use that to tether your two samples. I don't know who you
link |
are. I know nothing about you. I only know when those samples came out of a person,
link |
but I can say, oh, these two samples a year apart actually belong to the same person.
link |
Yeah, so there's sufficient information in that immunological history to match the samples.
link |
Or from a privacy perspective, that's really exciting. Does that generally hold for humans?
link |
So you're saying there's enough uniqueness to match? Yeah, because it's very stochastic,
link |
even twins. So this, I believe, we haven't published this yet. We will soon.
link |
You have a twin too, right? I do have a twin. I have an identical twin brother,
link |
which makes me interested in this. He looks very much like me. Oh, is that how that works?
link |
And DNA can't really tell us apart. But this tool is one of the only tools in the world that
link |
could tell twins apart from each other, could still be accurate enough to say this blood,
link |
it's like 99.999% accurate to say that these two blood samples came from the same individual.
link |
And it's because it's a combination both of your immunological history,
link |
but also how your unique body responds to a pathogen, which is random. The way that we
link |
make antibodies is by and large, it's got an element of randomness to it. How the cells,
link |
when they make an antibody, they chop up the genetic code to say, okay, this is the antibody
link |
that I'm going to form for this pathogen. And you might form, if you get a coronavirus, for
link |
example, you might form hundreds of different antibodies, not just one antibody against the
link |
spike protein, but hundreds of different antibodies against all different parts of the virus.
link |
So that gives this really rich resolution of information, that when I then do the same thing
link |
across hundreds of different pathogens, some of which you've seen, some of which you haven't,
link |
it gives you an exceedingly unique fingerprint that is sufficiently stable over years and years
link |
and years to essentially give you a barcode. And I don't have to know who you are,
link |
but I can know that these two specimens came from the same person somewhere out in the world.
link |
It's so fascinating that there's this trace, your life story in the space of viruses,
link |
in the space of pathogen. Because there's this entire universe of these organisms that are trying to
link |
destroy each other. And then your little trajectory through that space leaves a trace. And then you
link |
can look at that trace. That's fascinating. And that data period is just fascinating. And the
link |
vision of making that data universally connected to where you can make, like infer things, and just
link |
like with the weather is really fascinating. And there's probably artificial intelligence
link |
applications there, start making predictions, start finding patterns.
link |
Exactly. We're doing a lot of that already. And that's how we had this going. I've been
link |
trying to get this funded for years now. And I've spoken to governments, everyone says,
link |
cool idea, not going to do it. Why do we need it? Oh, really? The why do you need it?
link |
The why do you need it? And of course now, I've wrote in 2015 about this,
link |
why this would be useful. And of course, now we're seeing why it would be useful.
link |
Had we had this up and running in 2019, had we had it going, we were getting blood samples from
link |
hospitals and clinics and blood donors from New York City, let's just say. We didn't run the
link |
first PCR test for coronavirus until probably a month and a half or two months after the virus
link |
started transmitting in New York City. So it's like with the rain, we didn't start wearing umbrellas
link |
or taking out umbrellas. Exactly, for two months. It's getting wet.
link |
But different than the rain, we couldn't actually see that it was spreading.
link |
And so Andrew Cuomo had no choice but to leave the city open. They were hints that maybe the
link |
virus was spreading in New York City, but he didn't have any data to back it up. No data.
link |
And so it was just week on week on week. And he didn't have any information to really go by to
link |
allow him to have the firepower to say, we're closing down the city. This is an emergency,
link |
we have to stop spread before it starts. And so they waited until the first PCR test
link |
were coming about. And then the moment they started running PCR tests, they find out it's
link |
everywhere. And so that was a disaster because of course New York City, it was just hit so bad
link |
because nobody was, we were blind to it. We didn't have to be blind to it. And the nice thing about
link |
this technology is we wouldn't have, with the exact same technology we had in 2017,
link |
we could have detected this novel coronavirus spreading in New York City in 2020,
link |
not because we changed, not because we are actually actively looking for this novel coronavirus,
link |
but because we would see, we would have seen patterns in people's immune responses using AI
link |
or just frankly using our just the raw data itself. We could have said, hey, it looks like there's
link |
something that looks like known coronavirus is spreading in New York, but there's gaps.
link |
You know, there's, for some reason, people aren't developing an immune response to this
link |
coronavirus that seems to be spreading to these normal things that, you know, and it just looks,
link |
the profile looks different. And we could have seen that and immediately, especially since we
link |
had an idea that there was a novel coronavirus circulating in the world, we could have very
link |
quickly and easily seen, hey, clearly we're seeing a spike of something that looks like a known
link |
coronavirus, but people are responding weirdly to it. Our AI algorithms would have picked it up
link |
and just our basic heck, you could have put it in an Excel spreadsheet, we would have seen it.
link |
Some basic visualization would have shown that. Exactly. We would have seen spikes and they would
link |
have been kind of like off, you know, immune responses that the shape of them just looked
link |
a little bit different, but they would have been growing and we would have seen it and it could
link |
have saved tens of thousands of lives in New York City. So to me, the fascinating question,
link |
everything we've talked about, so both the huge collection of data at scale,
link |
which is super exciting. And then the kind of obvious at scale solution to the current virus
link |
and future ones is the rapid testing. Can we talk about the future of viruses that might be
link |
threatening the, our very existence? So do you think like a future natural virus can
link |
have an order of magnitude greater effect on human civilization than anything we've ever seen?
link |
So something that either kills all humans or kills, I don't know, 60, 70% of humans. So some
link |
like something, something we can't even imagine. Is that, is that something that you think is
link |
possible? Because it seems to have not happened yet. So maybe like the entirety, whoever, whoever
link |
the programmer is of the simulation that sort of launched the evolution for the Big Bang,
link |
seems to not want to destroy us humans. Or maybe that's a natural side effect of the
link |
evolutionary process that humans are useful. But do you think it's possible that the evolutionary
link |
process will produce a virus that will kill all humans? I think it could. I don't think it's likely.
link |
And the reason I don't think it's likely is, well, on the one hand, it hasn't happened yet,
link |
in part because mobility is a recent phenomena. People weren't particularly mobile until
link |
fairly recently. Now, of course, now that we have people flying back and forth across the globe all
link |
the time, the chances of global pandemics has escalated exponentially, of course. And so
link |
on the one hand, that's part of why it hasn't happened yet. We can look at things like Ebola.
link |
Now, Ebola, we don't, we haven't generally had major Ebola epidemics in the past, not because
link |
Ebola wasn't transmitting and infecting humans, but because they were, it was largely affecting
link |
and infecting humans in disconnected communities. So you see in rural parts of Africa, for example,
link |
in Western Africa, you might end up having isolated Ebola outbreaks, but there weren't
link |
connections that were fast enough that would allow people to then spread it
link |
into the cities. Of course, we saw back in 2014, 15 massive Ebola outbreak that wasn't
link |
because it was a new strain of Ebola, but it was because there's new inroads and connections
link |
between the communities and people got it to the city. And so we saw it start to spread.
link |
So that should be a little bit foreshadowing of what's to come. And now we have this pandemic,
link |
we had 2009, we have this. There is a benefit, or there is sort of a natural check. And this is
link |
like kind of Latke Voltaire predator prey dynamic kind of systems, ecological systems and mathematics
link |
that if you have something that's so deadly, people will respond more, maybe with a greater panic,
link |
a greater sense of panic, which alone could destroy humanity. But at the same time,
link |
like we now know that we can lock down. We know that that's possible. And so if this
link |
was a worse virus that was actually killing 60% of people who was infecting, we would lock down
link |
very quickly. My biggest fear though is let's say that was happening. You need serious lockdowns if
link |
you're going to keep things going. So the only reason we were able to keep things going during
link |
our lockdowns is because it wasn't so bad that we were still able to have people work in the
link |
grocery stores. Still people work in the shipping to get the food onto the shelves.
link |
So on the one hand, we could probably figure out how to stop the virus,
link |
but can we stop the virus without starving? And I'm not sure that that if this was
link |
another acute respiratory virus that say it transmitted the same way, but say it actually
link |
did worse damage to your heart. But it was like a month later that people started having heart attacks
link |
in mass. It's like not just one offs, but really severe. Well, that could be a serious problem for
link |
humanity. So in some ways, I think that there are lots of ways that we could end up dying at the
link |
hand of a virus. I mean, we're already seeing it. I mean, my fear is still, I think coronaviruses
link |
have demonstrated a keen ability to destroy or to create outbreaks that can potentially be deadly
link |
to large numbers of people. Flu strains, though, are still by and large my concern.
link |
So you think the bad one might come from the flu, the influenza?
link |
Yeah, their replication cycle, they're able to genetically recombine in a way that coronaviruses
link |
aren't. They have segmented genomes, which means that they can just swap out whole parts of their
link |
genomes, no problem, repackage them, and then boom, you have a whole antigenic shift, not a drift.
link |
What that means is that on any occasion, any day of the year, you can have, boom, a new,
link |
whole new virus that didn't exist yesterday. And now with farming and industrial livestock,
link |
we're seeing animals and humans come into contact much more. Just the opportunities for an
link |
influenza strain that is unique and deadly to humans increases all the while transmission and
link |
mobility has increased. It's just a matter of time, in my opinion.
link |
What about from an immunology perspective of the idea of engineering a virus? So not just
link |
the virus leaking from a lab or something, but actually being able to understand the protein,
link |
like the everything about what makes a virus enough to be able to figure out ways to
link |
maybe targeted or untargeted attack by...
link |
It's the birth immunity.
link |
Yeah. Is that something, obviously that's somewhere on the list of concerns,
link |
but is that anywhere close of the top 10 highlights along with nuclear weapons and so on
link |
that we should be worried about? Or is the natural pandemic really the one that's much
link |
great a concern? I would say that the former, that manmade viruses and genetically engineered
link |
viruses should be right up there with the greatest concerns for humanity right now.
link |
We know that the tools for better or worse, the tools for creating a virus are there.
link |
We can do it. I mean, heck, the human species is no longer vaccinated against smallpox.
link |
I didn't get a smallpox vaccine. You didn't get a smallpox vaccine, at least I don't think.
link |
So if somebody wanted to make smallpox and distribute it to the world in some way,
link |
it could be exceedingly deadly and detrimental to humans and that's not even using your imagination
link |
to create a new virus. That's one that we already have. Unlike the past, when smallpox would
link |
circulate, you had large fractions of the community that was already immune to it.
link |
And so it wouldn't spread or it would spread a little bit slower, but now we have essentially,
link |
in a few years, we'll have a whole global population that is susceptible. Let's look at measles.
link |
We have an entire, I mean, measles. There are some researchers in the world right now which,
link |
for various reasons, are working on creating a measles strain that evades immunity. It's not
link |
for bioterrorism, at least that's not the expectation. It's for using measles as an
link |
oncolytic virus to kill cancer. And the only way you can really do that is if your immune system
link |
doesn't, you know, if you take a measles virus and there's, you know, we don't have to go into
link |
the details of why it would work, but it could work. Measles likes to target potentially cancer
link |
cells. But to get your immune system not to kill off the virus, if you're trying to use the virus
link |
to target it, you maybe want to make it blind to the immune system. But now imagine we took some
link |
virus like measles, which has an R0 of 18, transmits extremely quickly. And now we have
link |
essentially, let's say we had a whole human race that is susceptible to measles. And this is a virus
link |
that spreads orders of magnitude easier than this current virus. Imagine if you were to plug something
link |
toxic or detrimental into that virus and release it to the world.
link |
So it's possible to be both accidental and intentional? Absolutely. Yeah, an accident. So
link |
Mark Lipsitch is a good colleague of mine at Harvard. We're both in the, he's the director
link |
of the Center for Communicable Disease Dynamics from a faculty member. He's spoken very, very
link |
forcefully and he's very outspoken about the dangers of gain of function testing,
link |
where in the lab we are intentionally creating viruses that are exceedingly deadly under the
link |
auspices of trying to learn about them. So that if the idea is that if we kind of accelerate
link |
evolution and make these really deadly viruses in the lab, we can be prepared for if that virus
link |
ever comes about naturally or through unnatural means. The concern though is, okay, that's one
link |
thing, but what if that virus got out on somebody's shoe? Just what if? If the effects of an accident
link |
are potentially catastrophic, is it worth taking the chances just to be prepared a little bit for
link |
something that may or may not ever actually develop? And so it's a serious ethical quandary
link |
we're in. How to both be prepared, but also not cause a catastrophic mistake.
link |
As a small tangent, there's a recent really exciting breakthrough of Alpha 2,
link |
of Alpha Fold 2 solving protein folding or achieving state of the art performance on
link |
protein folding. And then I thought, proteins have a lot to do with viruses. It seems like
link |
being able to use machine learning to design proteins that achieve certain kinds of functions
link |
will naturally allow you to use, maybe down the line, not yet, but allow you to use machine
link |
learning to design basically viruses, maybe like measles for good, which is like to attack
link |
cancer cells, but also for bad. Is that a crazy thought or is this a natural place where this
link |
technology may go? I suppose all technologies can, which is for good and for bad. Do you think
link |
about the role of machine learning in this? Oh yeah, absolutely. Alpha Fold is amazing.
link |
It's an amazing algorithm, series of algorithms. And it does demonstrate, to me, it demonstrates
link |
just how powerful everything in the world has rules. We just don't know the rules. We often
link |
don't know them, but our brain has rules, how it works. Everything is plus and minus. There's
link |
nothing in the world that's really not at its most basic level, positive, negative. Obviously,
link |
it's all just charge. And that means everything. You can figure it out with enough computational
link |
power and enough. In this case, machine learning and AI is just one way to learn rules. It's an
link |
empirical way to learn rules, but it's a profoundly powerful way. And certainly, now that we are
link |
getting to a point where we can take a protein and know how it folds, given its sequence,
link |
we can reverse engineer that and we can say, okay, we want a protein to fold this way. What
link |
does the sequence need to be? We haven't done that yet so much, but it's just the next iteration of
link |
all of this. So let's say somebody wants to develop a virus. It's going to start with somebody wanting
link |
to develop a virus to defeat cancer, something good. And so it will start with a lot of money
link |
from the federal government for all the positives that will come out of it.
link |
But we have to be really careful because that will come about. There's no doubt in my mind
link |
that we will develop, we're already doing it. We engineer molecules all the time
link |
for specific uses. Oftentimes, we take them from nature and then tweak them.
link |
But now we can supercharge it. We can accelerate the pace of discovery. To not have it just be
link |
discovery, we have it be true ground up engineering. Let's say you're trying to make a new molecule
link |
to stabilize somebody with some retinal disease. So we come up with some molecule that can
link |
improve the stability of somebody with retinal degeneration. Just a small tweak to that to
link |
say make a virus that causes the human race to become blind. It sounds really conspiracy
link |
theoryish, but it's not. We're learning so much about biology and there's always an
link |
affair. Heck, look at how AI and just Google searches, they are every single day being leveraged
link |
by nefarious actors to take advantage of people, to steal money, to do whatever it might be,
link |
eventually probably to create wars or already to create wars. And I don't think there's any
link |
question at this point behind disinformation campaigns. And so it's being leveraged, this
link |
thing that could be wholly good, it's always going to be leveraged for bad. And so how do
link |
you balance that as a species? I'm not quite sure. The hope is, as you mentioned previously,
link |
that there's some that we were able to also develop defense mechanisms. And there's something
link |
about the human species that seems to keep coming up with ways to just like on the deadline,
link |
just at the last moment, figuring out how to avoid destruction. I think I'm like eternally
link |
optimistic about the human race not destroying ourselves, but you could do a lot of things
link |
that would be very painful. Yes. What we're doing it already, just, I mean, we are seeing how our
link |
regulation today. Right. We did this thing, it started as a good thing, regulation of medical
link |
products. But now it is unwillingly and unintentionally harming us. Our regulatory landscape,
link |
which was developed wholly for good in our country, is getting in the way of us deploying a tool
link |
that could stop our economies from having to be sputteringly closed, that could stop deaths
link |
from happening at the rate that they are. I think we will come to a solution. Of course,
link |
now we're going to get the vaccine and it's going to make people lose track of why we even
link |
bother testing, which is a bad idea. But we're already seeing that we have this amazing capacity
link |
to both do damage when we don't intend to do damage and then also to pull up when we need to
link |
pull up and stop complete catastrophe. And so we are an interesting species in that way,
link |
that's for sure. So there's a lot of young folks undergrads. Grads, they're also young.
link |
Listen to this. So you've talked about a lot of fascinating stuff that's like, there's
link |
ways that things are done and there's actual solutions and they're not always intersecting.
link |
Do you have advice for undergraduate students or graduate students or even people in high school
link |
now about life, about career, how they might be able to solve real big problems in the world,
link |
how they should live their life in order to have a chance to solve big problems in the world?
link |
It's hard. I struggle a little bit sometimes to give advice because the advice that I give
link |
from my own personal experience is necessarily distinct from the advice that would make other
link |
people successful. I have unending ambitions to make things better. And I don't see barricades
link |
where other people sometimes see barricades. Now, even just little things, like when this
link |
virus started, I'm a medical director at Brigham and Women's Hospital. And so I oversee or helped
link |
oversee molecular virology diagnostics. So when this virus started, wearing my epidemiology hat
link |
and wearing my viral outbreak hat, I recognized that this is going to be a big virus that was
link |
important at a global level. Even if the CDC and WHO weren't ready to admit that it was a pandemic,
link |
it was obvious in January that it was a pandemic. So I started trying to get a test built at the
link |
Brigham, which is one of Harvard's teaching hospitals. The first encounters I had with
link |
the upper administration of the hospital were pretty much, no, why would we do that? That's
link |
silly. Who are you? And I said, well, okay, don't believe me, sure. But I kept pushing on it.
link |
And then eventually I got them to agree. It was really only a couple of weeks before the
link |
Biogen conference happened. We started building the test. I think they started looking abroad and
link |
saying, okay, this is happening. Sure. Maybe he was right. But then I went a step further and I said,
link |
we're not going to have enough tests at the hospital. And so my ambition was to get a better
link |
testing program started. And so I figured what better place to scale up testing than the Broad
link |
Institute? Broad Institute is amazing, very high throughput, high efficiency research institute
link |
that does a lot of genomic sequencing, things like that. So I went to the Broad and I said,
link |
hey, there's this coronavirus that's obviously going to impact our society greatly. Can we start
link |
modifying your high efficiency instruments and robots for coronavirus testing? Everyone in my
link |
orbit in the hospital world just said, that's ridiculous. How could you possibly plan to do
link |
that? It's impossible. And to me, it was the most dead simple thing to do. But the higher ups and
link |
the people who think about, I think one of the most important things is to recognize that most
link |
people in the world don't see solutions. They just see problems. And it's because it's an easy
link |
thing to do. Thinking of problems and how things will go wrong is really easy because you're not
link |
coming up with a brand new solution. And this to me was just a super simple solution. Hey,
link |
let's get the Broad to help build tests. Every single hospital director told me no,
link |
like it's impossible. My own superiors, the ones I report to in the hospital said, Mike,
link |
you're a new faculty member. Your ideas probably would be right, but you're too naive and young
link |
to know that it's impossible. And obviously, now the Broad is the highest throughput laboratory in
link |
the country. And so I think my recommendation to people is as much as possible, get out of the
link |
mode of thinking about things as problems. Sometimes you piss people off. I could probably
link |
use a better filter sometimes to try to be not so upfront with certain things. But it's just so
link |
crucial to always just see, to just bring it, like think about things in new ways that other
link |
people haven't because usually there's something else out there. And one of the things that has
link |
been most beneficial to me, which is that my education was really broad. It was engineering
link |
and physics. And well, and then I became a Buddhist monk for a while. And so that gave me a different
link |
perspective. But then it was medicine and immunology. And now I've brought all of it together from
link |
mathematics and biology and medicine, perspective, and policy and public health. And I think that
link |
I'm not the best in any one of these things. I recognize that there are going to be geniuses
link |
out there who are just worlds better than me at any one of these things that I try to work on.
link |
But my superpower is bringing them all together and just thinking, and that's,
link |
I think, how you can really change the world. I don't know that I'll ever change the world
link |
in the way that I hope. But that's how you can have a chance.
link |
Yeah, that's how you can have a chance, exactly. And I think it's also what,
link |
this to me, this rapid testing program, this is the most dead simple solution in the world.
link |
And this literally could change the world. It could change the world. And it is. There's
link |
countries that are doing it now. The US isn't, but I've been advising many countries on it.
link |
And I would say that some of the early papers that we put out earlier on, a lot of the things
link |
actually are changing. Unless you really look hard, you don't know where you're actually having
link |
an effect. Sometimes it's more overt than other times. In April, I published a paper that was
link |
saying, hey, with the PCR values from these tests, we need to really focus on the CT values,
link |
the actual quantitative values of these lab based PCR tests. At the time, all the physicians
link |
and laboratory directors told me that was stupid. Why would you do that? They're not
link |
accurate enough. And of course, now it's headline news that Florida, they just mandated
link |
reporting out the CT values of these tests because there's a real utility of them.
link |
You can understand public health from it. You can understand better clinical management.
link |
And that was a simple solution to a pretty difficult problem. And it is changing the way
link |
that we approach all of the lab testing in this country is starting to, it's taken a few months,
link |
but it's starting to change because of that. And that was just me saying, hey, this is something
link |
we should be focusing on. Got some other people involved and other people. And now people recognize,
link |
hey, there's actual value in this number that comes out of these lab based PCR tests. So sometimes
link |
it does grow fairly quickly. But I think the real answer, my only answer, I don't know what,
link |
I recognize that everyone, some people are going to be really focused on and have one small,
link |
but deep skill set. I go the opposite direction. I try to bring things together.
link |
But the biggest thing I think is just don't see barriers. Just see,
link |
like, there's always a solution to a barrier. If there's a barrier that literally means a
link |
solution to it, that's why it's called a barrier. And just like you said, most people will just
link |
present to you, only be thinking about it and present to you with barriers. And so it's easy
link |
to start thinking that's all there is in this world. And just think big. I mean, God, there's
link |
nothing wrong with thinking big. Elon Musk thought big. And, you know, and then thinking big builds
link |
on itself, you know, you get a billion dollars from one big idea and then that allows you to
link |
make three new big ideas. And there's a hunger for it. If you think big and you communicate
link |
that vision with the world, all the most brilliant and like passionate people will just like,
link |
you'll attract them and they'll come to you. And then it makes your life actually really
link |
exciting. The people I've met at like Tesla and Neuralink, I mean, there's just like this fire
link |
in their eyes. They just love life. And it's amazing, I think, to be around those people.
link |
I have to ask you about what was the philosophy, the journey that took you to becoming a Buddhist
link |
monk? And what were, what did you learn about life? What did you take away from that experience?
link |
How did you return back to Harvard and the world that's unlike that experience, I imagine?
link |
Yeah. Well, I was at Dartmouth at the time.
link |
Well, I went to Sri Lanka. I was already pretty interested in developing countries and sort
link |
of under resourced areas. And I was doing a lot of engineering work and I went there,
link |
but I was also starting to think maybe health was something of interest.
link |
And so I went to Sri Lanka because I had a long interest in Buddhism as well,
link |
just kind of interested in it as a thing. Which aspect of the philosophy attracted you?
link |
I would say that the thing that interested me most was really this idea of kind of a butterfly
link |
effect of what you do now has ripple effects that extend out beyond what you can possibly imagine,
link |
both in your own life and in other people's lives. And in some ways, Buddhism has,
link |
not in some ways, in a pretty deep way. Buddhism has that as part of its underlying
link |
philosophy in terms of rebirth and sort of your actions today propagate to others,
link |
but also propagate to sort of what might happen in your circle of what's called samsara and
link |
rebirth. And I don't know that I subscribe fully to this idea that we are reborn,
link |
which always was a little bit of a debate internally, I suppose when I was a monk.
link |
But it has always been, it was that and then it was also meditation. At the time I was a fairly
link |
elite rower. I was rowing at the national level and rowing to me was very meditative. It was
link |
just, there's, even if you're in a boat with other people, it's, I mean, on the one hand,
link |
it's like the extreme of like a team sport, but it's also the extreme sort of focus and
link |
concentration that requires, that's required of it. And so I was always really into just
link |
meditative type of things. I was doing a lot of pottery too, which was also very meditative.
link |
And so Buddhism just kind of really, really, there are a lot of things about meditating
link |
that just appealed. And so I moved to Sri Lanka planning to only be there for a couple of months.
link |
But then I was shadowing in this medical clinic and there was this physician who was just really,
link |
I mean, it's just kind of a horrible situation. Frankly, this guy was trained
link |
decades earlier. He was an older physician and he was still just practicing like these
link |
fairly barbaric approaches to medicine because he had, you know, as a rural town
link |
and he just didn't have a lot of, he didn't have any updated training, frankly. And so,
link |
you know, I just remember this like girl came in with like shrapnel in her hand and his solution
link |
was to like air it out. And so he was like, without even numbing her hand, he was like cutting it
link |
open more with this idea that like the more oxygen and stuff, you know, and it just, I think
link |
there was something about all of this. And I was already talking to these monks at the time.
link |
Each, I would be in this clinic in the morning and I'd go and my idea was to teach English to
link |
these monks in the evening. Turned out I'm a really bad English teacher. So they just taught,
link |
they allowed me just to sit with them and meditate and they were teaching me more about
link |
Buddhism than I could have possibly taught them about English or being an American or something.
link |
And so I just slowly, I just couldn't take, I like couldn't handle being in that clinic. So
link |
more and more, I just started moving to, you know, spending more and more time at this monastery.
link |
And then after about two months, I was supposed to come back to the States and I decided I didn't
link |
want to. So I moved to this monastery in the mountains primarily because I didn't have the
link |
money to like just keep living. So I was living in a monastery, it's free. And so I moved there
link |
and just started meditating more and more. And then months went by and it just really gravitated.
link |
I gravitated to the whole notion of it. I mean, it became, it sounds strange, but meditating almost,
link |
just like anything that you put your mind to, became exciting. It became like there weren't
link |
enough hours in the day to meditate. And I would do it for 18 hours a day, 15 hours a day,
link |
just sit there and you, and like, I mean, I hate sleeping anyway, but I wouldn't want to go to
link |
sleep because I felt like I didn't accomplish what I needed to accomplish in meditation that day,
link |
which is so strange because there is no end, you know, but it was always, but there are these,
link |
there are these steps that happen during meditation that are very prescribed in a way.
link |
Buddha talked about them, you know, and these are ancient writings, which exist. I mean,
link |
the writings are real. They're thousands of years old now. And, you know, so whether it was Buddha
link |
writing them or whoever, you know, there are lots of different people who have contributed to the,
link |
to these writings over the years. And, but they're very prescribed. And they, they tell you what
link |
you're going to go through. And I didn't really focus too much on them. I read a little bit about
link |
them, but your mind really does when you actually start meditating at that level, like not an hour
link |
here and there, but like truly just spending your day is meditating. It becomes kind of like this
link |
other world where it becomes exciting. And, and you're actively working, you're actively
link |
meditating, not just kind of trying to quiet things. That's sort of just the first stage
link |
of trying to get your mind to focus. Most people never get past that first stage,
link |
especially in our culture. Could you briefly summarize what's waiting beyond the stage of
link |
just quieting the mind? It's hard for me to imagine that there's something that could be described
link |
as exciting on there. Yeah, it's, it's an interesting question. So I would say,
link |
so the first thing, the first step is truly just to like be able to close your eyes, focus on your
link |
breath and not have other thoughts enter into your mind. That alone is just so hard to do.
link |
Like, I couldn't do it now if I wanted, but I could then. And, but once you get past that stage,
link |
you start entering into like all these other, you go through the kind of, I went through this
link |
like pretty trippy stage, which was a little bit euphoric, where you just kind of start not
link |
hallucinating. I mean, it wasn't like some crazy thing that would happen in a movie where you,
link |
but definitely just weird. You start getting to the stage where you, you're able to quiet your
link |
mind for so long, for hours at a time that like for me, I started getting really excited about
link |
this idea of mindfulness, which is part of, part of Buddhism in general, but it's part of
link |
Taravanan Buddhism in particular for this, in this way, which was, you take, you start focusing
link |
on your daily activities, whether that's sipping a cup of tea or walking or, you know, sweeping
link |
around. I lived in, on this mountain side in this cottage thing is built into the rock and,
link |
you know, so every morning I would wake up early and sweep around it and stuff, because that's
link |
just what we did. And you start to, you meditate on all those activities. And one of the things
link |
that was so exciting, which sounds completely ridiculous now, was just almost learning about
link |
your daily activities in ways that you never would have thought about before. So what is,
link |
what's, what's involved with like picking up this glass of water? You know, if I said,
link |
okay, I'm just going to pick, I'm going to take a drink of water, to me right now, it's a single
link |
activity. But during meditation, it's not a single activity. It's a whole series of
link |
activities of like little engineering feats and feelings. And it's gripping the water.
link |
And it's feeling that the glass is cold and it's lifting and it's moving and dragging and dragging.
link |
And you start to learn a whole new language of life. And that to me was like this really
link |
exhilarating thing that it was an exhilarating component of meditation that there was never
link |
enough time, it's kind of like learning a new computer language, like it gets really exciting
link |
when you start coding and all these new things you can do. You learn how to much, to experience
link |
life in a much richer way. And so you never run out of ways to go deeper and deeper and
link |
deeper in the way you experienced, even just the drinking of the glass of water.
link |
That's exactly right. And what becomes kind of exhilarating is you start to be able to predict
link |
things that you never, or I don't even have predictions or a word, but I always think of
link |
the matrix where I forget who it was, somebody was shooting at Neo and he like leans backwards
link |
and he dodges the bullets. In some ways, when you start breaking every little action that
link |
your hands do or that your feet do or that your body does down into all these little actions
link |
that make up one what we normally think of as an action, all of a sudden you can start to see
link |
things almost in slow motion. I like to think of it very much like language. The first time
link |
somebody hears a foreign language, it sounds really fast usually. You don't hear the spaces
link |
between words. And it just sounds like a stream of consciousness, and it just sounds like a
link |
stream of noises if you've never heard the language before. And as you learn the language,
link |
you hear clear breaks between words and it starts to gain context and all of a sudden
link |
like that, what once sounded very fast slows down and it has meaning. That's our whole life.
link |
There's this whole language happening that we don't speak generally. But if you start to speak it
link |
and if you start to learn it and you start to say, hey, I'm picking up this glass is actually 18
link |
little movements, then all of a sudden it becomes extremely exciting and exhilarating to just
link |
just breathe. Breathing alone in the rise and fall of your abdomen or the way the air pushes in
link |
and out of your nose becomes almost interesting. And what's really neat is the world just starts
link |
slowing down. And I'll never forget that feeling. And if there was one euphoric feeling from meditation
link |
I want to gain back, but I don't think I could without really meditating like that again,
link |
and I don't think I will. Was this like slow motion of the world? It was finding the spaces
link |
between all the movements in the same way that the spaces between all the words happen.
link |
And then it almost gives you this new appreciation for everything. It was really amazing. And so I
link |
think it came to an abrupt end though when the tsunami hit. I was there in the Indian Ocean tsunami
link |
hit in 2004. And it was like this dichotomy of being a monk and just meditating in this extraordinary
link |
place. And then the tsunami hits and kills 40,000 people in a few minutes on the coast of this really
link |
small little country in Sri Lanka. And then I like my whole world of being a monk came crashing down
link |
on when I go to the coast. And I mean, that was just a devastating visual site, an emotional site.
link |
But the strangest thing happened, which was that everyone just wanted me to stay as a monk.
link |
People in that culture, they wanted to, the monks largely fled from the coastlines.
link |
And so then there I was and people wanted me to be a monk. They wanted me to stay on the coast,
link |
but be a monk and not help. Not help in the way that I considered helping. They wanted me just
link |
to keep meditating so that they could bring me offerings and have their karmic responsibilities
link |
attended to as well. And so that was really bizarre to me. It was like, how could I possibly just sit
link |
around while all these people, half of everyone's family just died. And so in any case, I stopped
link |
being a monk and I moved to this refugee camp and lived there for another six months or so and
link |
just stayed there, not as a monk, but tried to raise some money from the U.S. and tried to,
link |
like, I didn't know what I was doing. Frankly, I was 22. And I don't think I appreciated at the
link |
time how much of a role I was having in that community's life. But it's taken me many years
link |
to process all of this since then. But I would say it's what put me into the public health world,
link |
seeing it living in that refugee camp and that difference that happened from being a monk to
link |
being in this devastating environment just really changed my whole view of what sort of why I was
link |
existing, I suppose. Well, so there's this richness of life in a single drink of water
link |
that you experience. And then there's this power of nature that's capable to take lives of thousands
link |
of people. So given all that, the absurdity of that, let me ask you, and the fact that you study
link |
things that could kill the entirety of human civilization, what do you think is the meaning
link |
of this all? What do you think is the meaning of life, this whole orchestra with God going on?
link |
Does it have a meaning? And maybe from another perspective, how does one live a meaningful
link |
life if such as possible? Well, from what I've seen, I don't think there's a single answer to
link |
that by any stretch. One of the most interesting things about Buddhism to me is that the human
link |
existence is part of suffering, which is very different from Judeo Christian existence, which
link |
is that human existence is something very different. There's a richness to it. In Buddhism, it's just
link |
another one of your lives. But it's your opportunity to attain nirvana and become a monk, for example,
link |
and meditate to attain nirvana. Else, you just go back into the samsara, the cycle of suffering.
link |
In some ways, the notion of life and what the purpose of life is, they're completely distinct,
link |
this western view of life, which is that this life is the most precious thing in the world,
link |
versus this is just another opportunity to try to get out of life. The whole notion of nirvana
link |
and in Buddhism, getting out of this sort of cycle of suffering is to vanish. If you could
link |
attain nirvana throughout this life, the idea is that you don't get reborn.
link |
On the one hand, you have Christian faith and other things that want to go to heaven and live
link |
forever in heaven. Then you have this other whole half of humans who want nothing more than to
link |
get out of the cycle of rebirth and just not exist anymore. The cycle of suffering.
link |
Yeah. How do you reconcile those two? Do you have both of them in you? Do you basically
link |
oscillate back and forth? I think we're just a bunch of proteins that we've formed and they work
link |
in this really amazing way. They might work in a bigger scale. There might be some
link |
connections that we're not really clear about, but they're still biological. I believe that
link |
they're biological. How do these proteins become conscious and why do they want to
link |
help civilization by having at home rapid tests a scale? Well, I think I don't have an answer to
link |
that one, but I really do believe that it's just an evolution of consciousness. I don't personally
link |
think my feeling is that we're a bunch of pluses and minuses that have just gotten so complex that
link |
they're able to make rich feelings, rich emotions. I do believe though, on the one hand,
link |
I sometimes wake up some days, my fiance doesn't always love it, but I think we're all just a
link |
bunch of robots with pretty complicated algorithms that we deal with. In that sense,
link |
like, okay, if the world just blew up tomorrow and nothing existed the day after that,
link |
it's just another blip in the universe. But at the same time, I don't know. So that's
link |
kind of probably my most core basic feeling about life. It's like we're just a blip and we may as
link |
well make the most of it while we're here blipping. It's one hell of a fine blip though.
link |
It is. It's an amazing blink of an eye in time. Michael, you're one of the most
link |
interesting people I've met, one of the most interesting conversations, important ones.
link |
Now, I'm going to publish it very soon. I really appreciate taking the time. I know
link |
how busy you are. It was really fun. Thanks for talking to me. Well, thanks so much. This was
link |
a lot of fun. Thanks for listening to this conversation with Michael Mina and thank you to
link |
our sponsors. Brave, a fast browser that feels like Chrome, but has more privacy preserving
link |
features. Athletic Greens, the only one drink that I start every day with to cover all my
link |
nutritional bases. ExpressVPN, the VPN I've used for many years to protect my privacy on the internet
link |
and Cash App, the app I use to send money to friends. Please check out the sponsors in the
link |
description to get a discount and to support this podcast. If you enjoy this thing,
link |
subscribe on YouTube, review it with five stars on Apple Podcast, follow on Spotify,
link |
support on Patreon, or connect with me on Twitter at Lex Freedman. And now, let me leave you with
link |
some words from Teddy Roosevelt. It is not the critic who counts. Not the man who points out
link |
how the strong man stumbles or where the doer of deeds could have done them better. The credit
link |
belongs to the man who actually is in the arena, whose face is marred by dust and sweat and blood,
link |
who strives valiantly, who errs, who comes short again and again because there is no effort without
link |
error and shortcoming, but who does actually strive to do the deeds, who knows great enthusiasm,
link |
the great devotions, who spends himself in a worthy cause, who at the best knows in the end
link |
that triumph of high achievement and who at the worst, if he fails, at least fails while daring
link |
greatly, so that his place shall never be with those cold and timid souls who neither know victory
link |
nor defeat. Thank you for listening and hope to see you next time.