Jeffrey Barrett

Good to be realistic, but compare it to the US regime proposing massive cuts to all science funding, with the legislative process as the only hope to forestall it!

This is great news. >80% increase (though less in real terms, of course) compared to <20% increase between the previous 7 year period ("Horizon 2020" was €80B) and the current one. Exactly the right response to this moment, and now we're talking real numbers.

Really nice work. This kind of paper is so important.

"We built cyber security into every stage of the education process. We established Cyber First and cyber apprentices to make sure that we got the talent we needed coming into the field."

"cyber apprentices" 😂

ICYMI it's not a Max Planck directorship but still a great opportunity to join an exciting research environment down under 😁 bsky.app/profile/mike...

Who said “cyber” in 2015? What did he even mean by it?

“Garber, in a conversation with a faculty member, said the suggestion Harvard was open to paying $500 million is ‘false’ and claimed the figure was apparently leaked by White House officials…Garber said the University is treating academic freedom as nonnegotiable.”
www.thecrimson.com/article/2025...

Harvard is seriously considering to resolve dispute "through the courts rather than a negotiated settlement"

"A deal with the Trump administration is not imminent"

Well OK then, Harvard holding strong after all! Let it be inspiration for other unis who have not folded yet 💪

Cleveland Museum photo showing an ancient Egyptian paint box carved from boxwood with five oval wells containing original paint pigments. From top to bottom the preserved cake pigments are red (red ochre), blue (Egyptian blue), green (a mixture of Egyptian blue, yellow ochre, and orpiment) and two of black (carbon black, from charcoal). Dimensions 2.2 cm x 21 cm x 3.6 cm. On the front of the box there’s an inscription inlaid in Egyptian blue with the name Amenemope who was Vizier during the reign of Amenhotep II, Dynasty 18. It translates as ‘The overseer of the city and vizier Amenemope. Master of secrets of the west bank [of Thebes]’. A red inscription on the back reads ‘Amen-Ra, Lord of the Thrones of the Two Lands’. Dated c. 1427-1401 BC.

Wow, this 3,400 year-old ancient Egyptian paint box still contains original pigments!

Looks similar to a modern-day set!

An inscription tells us it belonged to Amenemope, Vizier during the reign of Amenhotep II.

📷 Cleveland Museum of Art www.clevelandart.org/art/1914.680

#Archaeology

A few thoughts on Herasight, the new embryo selection company. First, their whitepaper (drive.google.com/file/d/1EpFi...) implies that competitors like Nucleus have been marketing and selling grossly erroneous risk estimates. This is shocking if true! 🧵

Oh, gross.

Text from an FAQ in Okbay et al 20222: https://www.nature.com/articles/s41588-022-01016-z a similar same statement is made in an FAQ in 2025: https://www.biorxiv.org/content/10.1101/2025.05.14.653986v1.supplementary-material Text reads: "The results of SSGAC studies have sometimes been used by online platforms, including some companies, to predict individual outcomes. We recognize that returning individual genomic “results” can be a fun way to engage people in research and other projects and to feed or stoke their interest in genomics. But it is important that participants/users understand that these individual results are not meaningful predictions and should be regarded essentially as entertainment. Failure to make this point clear risks sowing confusion and undermining trust in genetics research"

It is depressing, but all too predictable, how swiftly we’ve gone from the Social Science Genetic Association Consortium offering reassurances about the uses of behavioural polygenic scores to one of their lead authors marketing embryo selection for IQ

Absolutely brilliant piece by Ian Dunt. “You can have the life you want. And by pursuing it, you can make the world a better place. That seems a better story than the one of terror, insecurity and jealousy that we see play out online.”

True, python evaluates the expression the same way!

I knew it was a floating point rounding thing, but I wasn’t sure why this sum is affected. Turns out to be because 0.4 is a recurring decimal in binary, and so the product needs to be rounded.

Image of an old building in Oxford with the heading 'postdoc opportunities' and the text 'computational approaches to improve rare disease diagnosis and treatment' and 'Big Data Institute, University of Oxford'

📣 We are recruiting! Please share!!

Are you a bioinformatician / computational scientist who wants to apply your skills to understanding regulatory biology and improving rare disease diagnosis and treatment? 🧠 💻 🧬 🩺

We have two roles available 👇

🧵 1/4

Here's a fun bit of math in R:

Which way? Did you slam the door in someone’s face or stand there awkwardly waiting for them to catch up?

Someone needs to send this to the Harvard corporation every day

o3 does provide links to sources, but often can't coherently link a statement it makes to the relevant source. but yeah, this case was partially me phrasing the question ambiguously (because I'm unfamiliar) and it being super confident in the (wrong, somewhat goofy) interpretation

It takes my brain about 3 or 4 days to realise that the world will continue to turn without my active input; many sensible decisions can happen without my words of wisdom or - worse case - can wait for my return; and I can just ... let go. See y'all in 2 weeks' time!

It was, of course, absolutely convincing in its presentation of a 🍌 argument. Of course I know not to just trust AI's pronouncements, but this was the first time where it wasted my time compared to starting with good old Google. (and I'm sure sadly many people would just "ship" the first answer)

Just had a revealing experience with ChatGPT o3. I've previously found it to be a good first source to understand a science topic I'm not familiar with, and provide a jumping off point. But it just gave me a totally incorrect answer that sent me off in entirely the wrong direction for half an hour.

Yes, I was more careful and said "ChatGPT et al" in the first tweet, then was a bit sloppier in the second. I agree with the point!

I actually don't think it's the end of the world to embrace "AI = ChatGPT" in non-technical discourse, since it is so widespread. If you're making a technical distinction, just say what you mean. Anyone talking about a scientific/technical use case and just saying "we use AI" is already suspect.

I think "AI slop" is not a bad choice. It's too late to differentiate to a general audience about the type of model, etc -- to most people AI means ChatGPT et al., and adding "slop" makes it clear that you're disparaging it!

Etymology fact of the day: the Finnish word for "sale" is "ale", but the similarity is a coincidence! Rather than being an adopted germanic-root word, which happens often in Finnish (usually via Swedish), it's a shortening of "alennusmyynti" where "alennus" = discount from proto-Finnic "ala"=under.

PhD position for January start with me and Neil Ferguson, doing some neat model development for ethnicity-stratified transmission models. Please get in touch with any questions.

Home fees (UK) only unfortunately.

This implies there is a huge literature of small microbiome association studies that is JUST NOISE. (As observed for these other fields)

Was in Edinburgh at Easter, and spent a very pleasant afternoon walking around that area. Absolutely gorgeous city and university.