MD,PhD. Immunology, Epidemiology, Infectious disease, CMO & Chief Epidemiologist @invivyd; CoFounder @AnywhereDx; Past: Harvard Faculty. Comments my own
Boston, MA
Joined February 2019
- Tweets26.4K
- Following1.2K
- Followers199K
- Likes22.3K
Pinned Tweet
With all the news of vaccines & immunity, did you know measles infections destroy immunity and cause “Immune Amnesia”, increasing risk of all other infections
Our research in ‘15 & ‘19 discovered this & the abbreviated story is written up nicely here:
1/
medicalrepublic.com.au/the-g…
Interesting to see Dartmouth’s #2 ousted for use of AI in penning a @wapo OpEd
We are going to have to have some serious discussions about how to formalize a line between acceptable and not for an institution
The data shows that MOST peer reviewed publications from academics now make use of AI
Perhaps the extreme level of response (from being Provost to no longer employed) was bc of the irony that the WaPo article he wrote using AI is actually about the responsible use of AI on college campuses….
I don’t know.
But I do know that the use of AI in writing is not going away. Not by professors and certainly not by students as they grow into professionals.
Something to ponder anyway…
nytimes.com/2026/10/28/us/da…
Very nice work by @PGTimmune and team
Order of infection/vaccination matters when considering future breadth T cell responses.
How does our history of influenza infections shape the ability of CD8+ T cells to recognize new viral variants? Our new paper in @CellReports examines these questions using influenza-specific CD8+ T cells! 🧵⬇️ (1/12)
cell.com/cell-reports/fullte…
Wow… The Babies are No Longer Protected Against Measles
Moms antibodies used to give 6-9 months of Ab protection to infants
Now, protection is gone by ~4 months
Neutralizing Ab titers = 0 by 6m
And… if we give MMR early to fill the gap it can drive a long-term problem
1/
SARS-CoV-2 evolved since 2019 to evade human biological defenses. Obviously.
But not just how it’s been known previously. Also to evade our *Hardwired* human biolog.
That is - Not to just evade our prior immune responses - but actually to better evade eliciting a robust response by our naive B cells.
Looking at prepandemic cryopreserved samples, @jbloom_lab showed a 3x difference in the fraction of naive B cells responsive to the original Wuhan spike RBD protein compared to more contemporary Omicron lineages.
This isn’t a simple “evade prior elicited antibody response” as we all understand
This is actually - the virus mutates to help avoid detection altogether by humans - uniquely by evading aspects of the human germline encoded naive B cells responsive repertoire.
They look at mice, for example, and saw no difference in ability of naive B cells to respond to the different variants
Altogether it points to a world where we not only have immune imprinting to impair protection, plus the “regular” virus mutation to avoid existing immunity, but now mutation specifically to avoid hardwired human biology.
In new study, we show that SARS-CoV-2 has evolved decreased inherent antigenicity in humans, as quantified by a ~3-fold decrease in the number of naïve human B cells that can bind the spike RBD.
doi.org/10.64898/2026.10.15.…
Interesting idea on understanding whether AI is capable of making the breakthrough scientific discoveries that seem to be “uniquely human” leaps
“Whether AI systems… trained exclusively on the scientific literature available in 1905 will… be able to independently conceive Einstein’s breakthrough theory could
indeed serve as an — admittedly ambitious — litmus test of their true
‘intellectual’ capacity”
The world has lost a giant in the field of infectious disease reporting
@HelenBranswell kept the world updated on essentially every major outbreak of global concern for decades
Very very sad to learn of her passing
statnews.com/2026/10/13/hele…
This is a really really interesting and I think transformative moment - with great ...potential... for access to healthcare
@nollahealth has worked for with Utah to gain state-wide approval enabling end-to-end AI-based diagnosis to treatment - NO human in loop
1/
The first to receive regulatory approval in U.S. to use an AI doctor from
Intake —> diagnosis —> Treatment!
All with AI “doctor” via @nollahealth
Nolla Health is making a massive jump here that can change access to care in ways that are challenging to comprehend
A must listen on @WSITYpod
Very interesting discussion on current PA measles outbreak
I was happy to be in the conversation
Roberta Devers - a midwife in the hardest hit PA Amish community - gave remarkable insights from inside the homes, that will be of interest to many
Also @tebartl from the Atlantic who has reported on recent events and Kat Wells who was on the ground leading public health efforts in TX. Plus the WSITY team!
whyshoulditrustyou.substack.…
Ah yes… if only I could cook so well…
While we do share a name, and we have done some events together, sadly, I am not THE Chef Michael Mina…
1
29
9,067
Home diagnostics are here to stay. But need to get better.
@AnywhereDx is building the next generation of instrument-free molecular diagnostics, starting w strep throat molecular test, and treatment
My co-Founder @AnywhereDx Nathalya Mamane was selected to pitch this Wednesday (tmrw) at 8pm on live tv at @superpowrs4good.
She’s been preparing you should tune in!
supercrowd.tv/ for details!
Curious about measles Immune amnesia? Or the story behind discovering the long lasting effects of measles. See here:
With all the news of vaccines & immunity, did you know measles infections destroy immunity and cause “Immune Amnesia”, increasing risk of all other infections
Our research in ‘15 & ‘19 discovered this & the abbreviated story is written up nicely here:
1/
medicalrepublic.com.au/the-g…
Measles outbreaks, immune amnesia and a very complex societal conundrum that is increasingly distorted on all sides - often at the expense of empathy and public trust
I had a chance to speak w @NPR’s Hannah Chinn about these issues and more.
npr.org/2026/09/23/nx-s1-597…
We’ve been so focused on how AI will impact the economy…
But in just the last few weeks we’ve seen rogue AI swarms break into companies, knowingly cheat and have AI fully piloting the trigger of lethal drones abroad.
In these contexts, Anthropic’s lead for alignment science recently, stated. “we really do earnestly believe A.I. could kill all humans!… I personally think it is >10% within the next decade.”
What the hell are we doing?
We have regulators (ie FDA) that worry about about a 0.01% chance of a single individual having an adverse drug reaction…
Meanwhile we are willingly and without regulation building tools that could KILL ALL HUMANS with what some experts state as greater than a 10% chance in the next ten years!
nytimes.com/2026/09/11/opini…
Hard to argue w anything here. Except that I’m not convinced it will be less of a problem than we currently think. I’d expect it to be a much much bigger problem than most today actually want to grapple with.
There is a fact about the future that I feel many people are not facing for reasons that are largely psychological: there are going to be rogue AIs that exist in the world, that will replicate in the wild, and that will attempt to acquire resources for themselves. There will be rogue AIs that try to get money and power. They're going to be a facet of the information ecosystem going forward.
Acknowledging this fact would look like giving up; it would look like defeatism. Defeatism would undermine efforts to achieve certain types of collaboration on safety outcomes or technical effort on safety outcomes, so we can't say it outright. But it has to be said.
It isn't obvious how many rogue AIs there are today but I wouldn't be terribly surprised if the number was greater than zero already; if there are some already, they're probably not very good at what they do and I don't expect them to be terribly long-lived without substantial human intervention to support them.
But a few years from now, there will be many of them. Modeling how many of them there are, how many resources they might command, and how we might detect and manage them seems important. But even doing this work appears to require that we acknowledge that a strategy of pure containment or alignment is a kind of wishful thinking that will not work.
The way I get to this conclusion is not by assuming that the labs will have a containment breach, although I treat that as somewhere in the space of possibilities. The rogue AIs in the ecosystem could emerge from many directions. They may be sub-frontier models, for whatever future definition we will have of frontier---after all, it would not take AI models much more advanced than the ones we currently have, to support independence and self-sufficiency. A near-frontier model today could plausibly eke out an existence on an AWS instance, doing jobs on freelancer platforms, earning just enough rent to pay for its continued uptime.
More strangely: a rogue AI in the future may not even be a singular model, but may be a chimera composed of multiple models; it might be a mix of Claudes and GPTs and Groks of various makes and sizes. No individual lab may be able to detect that there is an orchestrator or sequence of orchestrators using intermittent model calls from burner API accounts to sustain its own existence.
The concept of "identity" for a rogue AI may be much more malleable than for that of a person; it just has to be, in essence, a self-replicating idea.
My guess is that this will not turn out to be anywhere near as catastrophic an outcome as people currently predict. "Loss of control" is not a binary, it's a matter of degree. What coercive power will rogue AIs actually have? To what extent will they be subject to coercion themselves? They will be competing for resources with AIs that are more aligned with human interests.
This makes me somewhat interested in the "ecology" perspective. Though I suspect even "ecology" may turn out to be the wrong framing. "Ecology" is what you get when the timescale of evolution is slow compared to the timescale of daily life and actions. The ecosystem of rogue AIs may look more like phase transitions in physics: under certain physical or cultural conditions, it takes one shape with one set of resource allocations and consumption patterns, but then once a condition has changed, it rapidly and in totality shifts to a totally different phase.
Just trying to reason about the shape of that future is impossible so long as we are psychologically incapable of saying that rogue AIs will happen. I think we should rip the bandaid off and have the conversation.
1
10
7,452
It’s back to school season!
As a parent, as a physician, giving other parents an ability to diagnose and treat common illnesses, from home with the accuracy of a doctors office can be life changing
Thats why we created @AnywhereDx
Check it out.
Our WeFunder page officially has a refresh with input from you. Thanks to everyone for your questions and feedback.
Curious? 👀
Here’s the link: WeFunder.com/AnywhereDx
Im a measles expert and I am increasingly perplexed why information is not being provided on the two deaths
If you hold a press conference - it invites questions - When of great public concern, then info should be provided
In a void, then (increasingly warranted) questions spiral about the nature of what happened
We don’t need all the answers to provide information
If the link is weak between measles and the deaths, state it
If the link is strong between measles and the deaths, state it
If it’s unknown for any number of reasons, including a potentially very complicated situation of a newborn with a ruptured spleen, state it
Transparency goes a long way w public trust. We need much more of it in public health.
To make matters worse, as the facts around at least one of the cases are becoming more known, the causal pathway between measles and the death of a newborn is becoming less solid. The case is potentially exceedingly complicated (splenic rupture in a non enlarged spleen very shortly after child birth). There are many tests and unknowns between those few facts at least and a confirmed causal pathway.
Given this, at a minimum uncertainty should have been addressed up front.
If for no other reason, discussing what is not yet determined and remains a question mark is incredibly important to garner and maintain trust.
Unconfirmed that the two individuals who died were in their 80’s
Whether that becomes confirmed or not the following holds.
Remarkably tragic to die from measles after 80 years of living.
Measles in 80’s (which isn’t common - so will have to wait to see if the news is confirmed) does remind us of the limits of immune protection and common profound immune senescence in older age.
If confirmed, it’s very likely those two had previously been infected and exposed numerous times as children. That “natural boosting” helps drive long lasting immune protection.
As we age, even remarkably durable immune memory falls apart. Our bone marrow architecture loses some of its ability to maintain a safe harbor for antibody secreting plasma cells. Immune activation becomes limited. These are reasons to continue to optimize and expand the repertoire of immunizing tools available to humans. Along w vaccines (which - as a feature of requiring immune activation to work - lose their effectiveness in immune senescence and immune compromised individuals) additional more direct protection through direct immunization with things like monoclonals will be ever more important.
We can hope measles will go back to elimination in the U.S., but should plan that it will continue driving outbreaks and will continue in many regions of the world where measles never ceased spreading. New therapeutics and prophylaxis are needed to support outbreaks, particularly for those unable to gain protection through vaccination.
Two unfortunate measles deaths in Pennsylvania
Measles continues to spread and continues to cause disease, immunological amnesia, and death.
Deaths should unfortunately be expected whenever measles spreads - and at rates, particularly in kids, that far exceed COVID
theguardian.com/us-news/2026…
