Richard Sever
richardsever.bsky.social
Richard Sever
@richardsever.bsky.social
Chief Science and Strategy Officer, openRxiv. Co-Founder, bioRxiv and medRxiv.
Reposted by Richard Sever
Getting Rosalind Franklin’s story right is crucial, because she has become a role model for women going into science

www.nature.com/articles/d41...
What Rosalind Franklin truly contributed to the discovery of DNA’s structure
Franklin was no victim in how the DNA double helix was solved. An overlooked letter and an unpublished news article, both written in 1953, reveal that she was an equal player.
www.nature.com
November 10, 2025 at 9:44 AM
Reposted by Richard Sever
I am seeing a lot of posts about Rosalind Franklin that themselves ignore her publication record on DNA!

In fact Franklin and Gosling's paper, including the famous Photograph #51, was published, along with Wilkins's paper, back-to-back with the Watson and Crick paper in Nature in 1953.
November 9, 2025 at 6:43 PM
Reposted by Richard Sever
If you believe either that Franklin discovered the double helix, and / or Watson and Crick stole her data, ask yourself how you know this. Then take a read of this article.
November 8, 2025 at 7:32 AM
"AI insight with human judgment".

After all the debate yesterday's announcement provoked, interesting to see Life Science Editors partner with qed - gets to the question of what the balance of human & machine review should look like. We'll see more of this.

www.lifescienceeditors.com/qed-pilot/
Limited Time AI offering! - Scientific editing and writing experts - Life Science Editors
AI that thinks critically. Editors who know what matters.
www.lifescienceeditors.com
November 8, 2025 at 4:31 PM
"Fig. 2 shows a representative experiment"
The science literature used to be more fun
November 8, 2025 at 3:49 PM
Great thread from Jeremy. But 16/41 leapt out and reminded me again of learning about base pairing in high school. That was _the_ moment I knew I wanted to study biochemistry.

Now I really must go read Heraclitean Fire...
He may not have fully understood this, but Crick did because of his work on helical diffraction. The question of where the bases were remained. According to The Double Helix, Watson was focused on this and eventually figured out the A-T and G-C pairing that proved so important.

16/41
November 8, 2025 at 2:14 PM
James Watson has died www.nytimes.com/2025/11/07/s...
James Watson, Co-Discoverer of the Structure of DNA, Is Dead at 97
www.nytimes.com
November 7, 2025 at 8:00 PM
And another thing: I am far more concerned with data provenance and information/identity verification than peer review right now… 1/2 journals.plos.org/plosbiology/...
November 7, 2025 at 2:30 PM
The reaction to this is fascinating. It's like a microcosm of all discussions around AI: lots of enthusiasm, lots of loathing (much of it reflexive), and some wise 'let's maybe try it and see' responses.
November 7, 2025 at 1:43 PM
Reposted by Richard Sever
Yep as a former editor who has seen thousands of human reviews, including one that was simply one word - "crap" - I concur that human performance levels in reviewing might not be what we should aspire to 🤣
November 7, 2025 at 1:32 PM
"We need more institutional experiments for science" makes sense to me @seemaychou.bsky.social. We might not like some of 'em and some'll fail, but given academic science often looks like a feudal system needing an industrial revolution trying things can't hurt.. seemay.substack.com/p/big-experi...
Big experiments are only big if they can fail
Some reflections on Arena Bioworks' unexpected wind down as a fellow institutional experimentalist
seemay.substack.com
November 6, 2025 at 7:08 PM
Reposted by Richard Sever
Please read the thoughtful Blog that @richardsever.bsky.social and the @openrxiv.bsky.social team wrote on integrating pre-prints with AI review openrxiv.org/enabling-rev...
November 6, 2025 at 3:11 PM
"Even with notable gains in fluency, GPT-5 is still prone to hallucinate, break rules, and exhibit latent capabilities that raise safety and biosecurity concerns" www.nature.com/articles/s41...
The fragile intelligence of GPT-5 in medicine
Nature Medicine - The latest large language model from OpenAI offers safety gains, persistent risks and the illusion of understanding.
www.nature.com
November 6, 2025 at 2:13 PM
Reposted by Richard Sever
Denmark eyes new law to protect citizens from AI deepfakes

Denmark plans to give citizens copyright over their likeness to ban and force removal of nonconsensual deepfakes (images, video or voice), targeting rising AI-generated abuse and misinformation.
Denmark eyes new law to protect citizens from AI deepfakes
Denmark plans to give citizens copyright over their likeness to ban and force removal of nonconsensual deepfakes (images, video or voice), targeting rising AI-generated abuse and misinformation.
apnews.com
November 6, 2025 at 11:08 AM
Reposted by Richard Sever
Bacteria can sense when a virus starts shredding their genome — by detecting methylated mononucleotides.
Here’s the story of how we discovered the Metis defense system 👇
www.biorxiv.org/content/10.1...
November 6, 2025 at 5:00 AM
Always great to see peer review experiments with actual data. SolvingForScience distinguish Quality (Q) & Impact (I) scores for papers. Q scores are based on a peer-improvement phase, followed by an I assessment by a different group. Unsurprisingly I varies more than Q 1/2
doi.org/10.1101/2025...
November 5, 2025 at 10:59 PM
"a true 'commons' for journal-agnostic preprint peer review" www.embo.org/features/pub...
Publishers join the Review Commons peer review process  – Features – EMBO
Editors from eLife, Genetics, G3 and Journal of Cell Biology will contribute to the EMBO preprint peer review platform
www.embo.org
November 5, 2025 at 10:17 PM
"Sustaining openness in an age of industrial-scale extraction may require new economic and governance models...it may be useful to consider historical precedent [like] governance of biological resources [from] indigenous communities" sverhulst.medium.com/the-weaponis...
The Weaponisation of Openness? Toward a New Social Contract for Data in the AI Era
By Stefaan G. Verhulst
sverhulst.medium.com
November 3, 2025 at 3:26 PM
THIS. In all the discussions about Publish-Review-Curate, alt peer review models, etc., too often the other things (good) journals (should) do are often omitted. Information verification and many other checks are as, if not more important, journals.plos.org/plosbiology/...
November 3, 2025 at 2:26 PM
Reposted by Richard Sever
AI doing by exactly what human recruitment panels do
This week's @nature.com cover highlights a report of A.I.- mediated distortion
When #ChatGPT was asked to rate 40,000 résumés, it ranked the older male candidates as better quality than the younger female applicants nature.com/articles/s41...
November 2, 2025 at 7:04 PM
I’ve spent many years explaining/defending biorxiv’s “no reviews” policy.

The logic was always that opening Word and opining is a far lower barrier than doing actual research, so noise’d be >> signal and we didn’t want to make subjective quality judgements.

LLMs mean it makes even more sense 1/2
arXiv will no longer accept review articles and position papers unless they have been accepted at a journal or a conference and complete successful peer review.

This is due to being overwhelmed by a hundreds of AI generated papers a month.

Yet another open submission process killed by LLMs.
Attention Authors: Updated Practice for Review Articles and Position Papers in arXiv CS Category – arXiv blog
blog.arxiv.org
November 2, 2025 at 5:33 PM
Reposted by Richard Sever
Where Parisians post their preprints…
October 31, 2025 at 6:28 PM
Reposted by Richard Sever
Vaccines have health benefits you didn’t even know about. Get your flu shots yearly, and shingles shots as soon as you are eligible.
Acute and chronic viral infections, including Herpes Zoster (Shingles), Hepatitis C, HIV, CMV, influenza, and SARS-CoV-2, are linked with a substantial increased risk of cardiovascular events, from a systematic review. Figure for SARS-CoV-2 below
newsroom.heart.org/news/some-ac...
October 29, 2025 at 1:44 PM
Scientific Reports meanwhile on course to publish 40K papers = >$100M pubmed.ncbi.nlm.nih.gov?term=%22Scie...
October 28, 2025 at 3:51 PM