Thomas House

New Preprint: "Symptom clusters in Long COVID in the UK: prospective community-based cohort study using unsupervised machine learning" Long COVID, alongside other post-viral conditions, remains a large problem with many open questions. We use a large community dataset to > arxiv.org/abs/2609.05213

He said we keep launching moral startups to beat capitalists to AGI, but progress just accelerates due to competition. So I asked where they got founders and he said the AI safety community and I said it sounds like you’re feeding AI safety researchers to capitalism and his daughter started crying.

I did a somewhat deep dive into the recent OpenAI and Anthropic "hacks" for an upcoming interview. IMO, the AI companies are better off running with the "rogue AI" story because the real story is actually pretty embarrassing. Like making your Wi-Fi password "wifi" embarrassing. 🧵

Great thread - I lived for some time in Coventry and while there was a big planning mistake made in designing it for the car, two tone was a famous example of a quite special civic culture.

The right way to pace the frontier is to commit to a robust, open-source AI ecosystem, thereby diluting the power of these badly behaving frontier labs.

THE PREMISE IS THAT THE MUPPETS ARE PUTTING ON A SHOW CALLED THE MUPPET SHOW

I'm calling it: Trump's dividend will be in Crypto, probably Trumpcoin.

S4 of #BiologyInNumbers begins with another set of minisodes recorded at the @smbmathbiology.bsky.social/ESMTB meeting. We got to visit the lovely city of Graz to talk about: women’s health, social determination of measles, and a precious baseball card collection... smb.org/news/13686145 #MathBio

The number of cases of cervical cancer reported in Scotland among women given the HPV vaccine at 12-13 years of age is now zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. Zero. www.bbc.com/news/article...

Only finally managed to see The Odyssey and yes it's too long and has many of Nolan's other weaknesses, but he's a tremendous cinematographer and his antipathy to CGI makes everything much scarier and more visceral - or in the case of the always amazing Samantha Morton, more magical.

Just remembering that the British police didn't feel so unable to act when it was women at a vigil ... www.bbc.co.uk/news/uk-5640...

This update is very welcome and will definitely reduce confusion.

What if light travelled super slow? Part 2 Things get even weirder with moving objects!

i find the demands that the media hound ariely to death as well horrifying. harassment shouldn't be the response to academic fraud for anyone. that it happened (at least) once is already one time too many.

Mark Kermode is, as usual, dead right here - "Teenage Sex and Death at Camp Miasma" is incredibly fun, Hannah Einbinder and Gillian Anderson are obviously having a massive blast with it and both do a great turn. www.youtube.com/watch?v=GZ76...

In the future, will the NHS use DNA tests to decide treatment plans? In the final episode of our latest podcast series we spoke to Joe Hilton, a health economist and @juniper-network.bsky.social Fellow, to understand how maths is used to define policies like these. plus.maths.org/joe-hilton-s...

#TodayinHistory #dataviz #OTD 📊 💀Major André-Louis Cholesky was killed in action during the First World War on 31st August 1918. His Cholesky algorithm iteratively simplified the solution of linear equations for regression coefficients. It served as inspiration for L.L. Thurstone's Factor Analysis

I have just discovered, although it's not surprising, that Google Maps will show you the territorial claims and names of whatever country you tell it you're from, dropping the "contested" boundaries you're probably used to seeing on World maps.

Although, in a world of limited characters, "nowt" can actually be quite useful.

New Preprint! "A framework for combined epidemiological-genomic inference to improve estimation of household model parameters" with @joebhilton.bsky.social and Golsa Sayyar. We extend the state space of standard household models to combine genomics with epidemiology. arxiv.org/abs/2608.21094

It seems that #Medicine at #Imperial College London was even more corrupt than indicated by the death of Stefan #Grimm: www.dcscience.net?s=Grimm Read about Kim Fox, open access in the BMJ, us.list-manage.com/F92wAAA8bpn?...

I just read in a history book that Swedish meatballs were Ottoman influenced, and of all the things to get caught up in a culture war ... (it may or may not be true or course - the book is by Baer whose general thesis is Europeans understate Ottoman influence). www.scmp.com/magazines/st...

Newborns rely on herd immunity because the MMR is not given til 12mths…. Incredibly sad

Lol.

$3.7 billion spent on APCs in 2025. That's ~1000X more than it costs to run bioRxiv/medRxiv. Worth thinking about as we consider how to fund infrastructure... arxiv.org/pdf/2608.16322

Meet thomrawson.bsky.social from oxforddemsci.bsky.social – another of our 2026 JUNIPER Fellows. Thom researches how demographic change affects disease transmission and healthcare demand, alongside real-time epidemic modelling and science communication. Congratulations, Thom!

RIP. Much to remember, but the advocacy and support for vaccines mentioned here is maybe lesser known.

Today we're kicking off a series celebrating our new Fellowship recipients! First up is Fanny @oxfordmathematics.bsky.social whose research uses statistical and computational modelling to support outbreak response and animal & public health policy, with a focus on avian influenza. #Epidemiology

Talking of which, which Nobel Laureate would have been a great poaster? I suspect Gell Mann would have been dryly devastating.

"When I called my book 'Leopardistan' I think I was pretty clear whose faces ought to be getting eaten" says journalist.

New Preprint! "A framework for combined epidemiological-genomic inference to improve estimation of household model parameters" with @joebhilton.bsky.social and Golsa Sayyar. We extend the state space of standard household models to combine genomics with epidemiology. arxiv.org/abs/2608.21094

Not all physicists of course, but a very good story about how not to be a helpful physical / mathematical scientist in biology!

One of the first "outrageous" claims I saw as Jason Arday's media hounding unfolded was that he ran 30 marathons in 35 days! Wow! What a thing to claim! And the newspapers all suggested there was no evidence he actually did it. Well, they clearly didn't look very hard 🙄 🧵 1/26

Anti-AI guy: People who find specific applications for this technology in their fields are lying and stupid. Pro-AI guy: Anyone who can’t find a use for this technology in their everyday life is lying and stupid. Enlightened synthesis: Everyone is lying and/or stupid.