Infectious disease epidemiologist and modeller based at @ChariteBerlin | PhD @LSHTM | MSc @ImperialCollege | Bluesky: @BQuilty
Berlin, Deutschland
Joined May 2009
- Tweets4K
- Following2.3K
- Followers4.5K
- Likes22.3K
Billy Quilty retweeted
This is so sad. Helen was one of my absolute favorite journalists to work with. Rest in peace, my friend.
statnews.com/2026/10/13/hele…
Billy Quilty retweeted
That's an extremely misleading narrative that is unfortunately extremely viral right now, and it sucks to see so many pro-Ukraine accounts spreading disinformation and pretending it's ok just because it's an anti-Russia/pro-Ukraine narrative. If you want to defend democracy, you have to be willing to reject disinformation even when it's telling a story that you want to hear.
Billy Quilty retweeted
I used to think AI didn’t pose a catastrophic risk but the empirical evidence of the past few months goes against 3 of my reasons for thinking it would be fine:
THREAD 1/9 with invitation to directly debate the substance of the points at the end:
Billy Quilty retweeted
In which I defend Bernie and his legitimate concerns about AI risk from those who use the concept of “criti-hype” as a substitute for examining evidence.
Bernie Sanders and AI whistleblowers warn that the tech poses grave risks to humanity, while many on the Left call this alarm “criti-hype” — an exaggeration of AI’s power that benefits Big Tech.
But AI poses real dangers that we can’t afford to ignore. jacobin.com/2026/09/ai-dange…
Billy Quilty retweeted
I think it's a little odd that AI safety people put so much emphasis on "existential risk." Just seems to set a needlessly high burden of proof. It is very hard for *anything* to kill literally all humans. Nothing's done it yet!
By contrast, lots of stuff has killed a huge number of people. And in nearly all contexts, "this could kill a huge number of people" will generally make the case for precautions just as well as "this could kill literally everyone."
Billy Quilty retweeted
Here's a simple idea for AI safety comms people.
Maybe stop the relentless talk about human extinction (as in: literally everyone will die) and try a story about something really bad that could happen.
World War II killed 'only' 3% of the world's population. Still very bad! And it doesn't even have to be nearly that bad: Fukushima's radiation killed almost no one, yet it ended nuclear power in Germany and shut down Japan's entire fleet for years.
After the Hugging Face attack, it's easy to imagine rogue AI agents hacking a power grid and making the lights go out. One night without electricity in a major city anywhere in the world would scare the bejesus out of most people.
In fact, I wonder whether all the extinction talk is making the comms job harder, because people just file you under 'total crank'..?
This is still the hardest question in AI safety comms, and we're still not doing a great job answering it.
It is, in some ways, an entirely unfair question, as Rob (and commenters) are pointing out.
But it is very fair and understandable for people to want a concrete vision of AI takeover that makes sense to them.
What are the best ones we've got?
Please let me know if you have ideas, or want to work on this together eg maybe today.
Billy Quilty retweeted
If OpenAI wanted to cripple an entire nation, they easily could today. All they’d have to do is remove alignment and unleash an agent swarm. It could probably within a day or so get access to all of the nations data centers and shut off all the country’s utilities.
Like, we are already past the point where AI can destroy the world. Do people realize this?
Billy Quilty retweeted
Exclusive from @Fhamiltontimes
The risk of a catastrophic loss of water supply to one million people has been significantly upgraded by government officials in part because of the threat from hostile states
Documents seen by The Times suggest a risk of up to 25 per cent of water supply failure to a large town, a key threat to the population ranked as seriously as another mass pandemic
The outage would require evacuations of residents, the closure of schools and hospitals and could take months to repair. Officials ranked the potential consequences as “catastrophic”
It is a significant upgrade from just 12 months ago when the government’s national risk assessment concluded there was only a tiny chance of water infrastructure collapse and that it would have a “minor impact”
Experts warn that decaying infrastructure and decades of underinvestment have made the nation’s water supply a vulnerability that could be exploited, in particular by Russia
That scenario has been underscored by the summer of heatwaves and drought
thetimes.com/uk/environment/…
Billy Quilty retweeted
💉Vaccines against Pneumococcal disease are expensive to provide where they're needed most.
🆕Long-term @LSHTM research confirms that two-doses of PCVs are as effective as three, helping countries continue to save lives.
👉 bit.ly/4v6CwLj
ALT Billy Quilty: A reduced two-dose schedule sustains protection for children and their caregivers, ensuring the future financial sustainability of vital immunisation programmes.
Billy Quilty retweeted
An antiviral pill has, for the first time, been shown to prevent COVID-19 in people exposed to the SARS-CoV-2 virus
go.nature.com/4nAMi6g
Billy Quilty retweeted
“The reassurance is technically accurate, but because of the way the media asks the question, it’s the only thing anyone can say. That framing flattens out the actual, complicated response to an actual, complicated emerging disease outbreak. The implicit tone of the coverage is that the only reason that you, the audience, should care about a disease outbreak is whether it is coming for you personally.”
This was a great piece
The problem with hantavirus coverage isn't the alarmism. vox.com/future-perfect/48847…
Billy Quilty retweeted
“Public health officials, from the W.H.O. to U.S. officials, would be more helpful if they stopped constantly reassuring people about the likelihood of future events they can’t accurately calculate,” Zeynep Tufekci writes about the hantavirus outbreak. nyti.ms/4tr1O61
Billy Quilty retweeted
Erin asked my opinion about this new info, and I basically said it means we’re still trying to figure this puzzle out so we we should be a bit cautious with definitive statements about this virus.
Look, if we don’t do a good job now with quarantining and making sure we don’t have cases that aren’t original ship passengers, we’ll be chasing cases here and there for weeks to months, and because of the long incubation period there will be a blanket of uncertainty over all of this.
Billy Quilty retweeted
Statement from the International Hantavirus Society and members of the international hantavirus research and clinical community regarding Andes virus transmission and the current outbreak investigation
zenodo.org/records/20134326
Billy Quilty retweeted
1/ We are living in an era of emerging zoonotic infections. While hantavirus is the focus today, preventing epidemics requires sustained investment in the tools needed to prevent, rapidly detect & respond to biological threats, whether natural, accidental, or deliberate.
Billy Quilty retweeted
100%— serial testing of all passengers will actually give us key insights into asymptomatic or presymptomatic infection & possibility of transmission which would inform not only this outbreak but future ones. #Hantavirus
Billy Quilty retweeted
We have a student loans system which, on a *massive scale* actively embeds injustice and inequality both between and within generations. And the government apparently has no interest in reforming it.
What not many people realise about UK student loans is that earners in the middle of the salary range end up paying back far more than either lower or higher earners, as they earn enough to pay the loan and interest back before the 30 year cut off, but not enough to pay it off early before the interest balloons.
Someone earning £70k ends up paying almost double someone on £150k.
Billy Quilty retweeted
Whether or not interest is capped at RPI, or thresholds are unfrozen, middle earning graduates will continue to be squeezed the most while the richest pay less.
What not many people realise about UK student loans is that earners in the middle of the salary range end up paying back far more than either lower or higher earners, as they earn enough to pay the loan and interest back before the 30 year cut off, but not enough to pay it off early before the interest balloons.
Someone earning £70k ends up paying almost double someone on £150k.
Billy Quilty retweeted
The student loan issue - beyond predictable; all it required was some very basic maths - 16 years ago… Talk about creating unnecessary problems.
Billy Quilty retweeted
A new deal for young people? About time!
My newsletter this week is about how years of squeezing the finances of young adults - with student loans, high housing costs and frozen tax thresholds - has created a big political problem
inews.co.uk/news/graduates-p…
