Oh, the irony of people using generative AI to "summarise" my papers 🫠 (here the real paper: link.springer.com/article/10.1...)

If you're interested but haven't registered, the virtual meeting is *very* cost effective, ranging from $10 for students to $40 for professional non-members! www.evolutionmeetings.org

New version of my gen AI academic slop graphics collection. bit.ly/academic-slop One of the new entries includes this 👇, which is part of the inner ear. 👂 Allegedly. H/T @mathchaos.bsky.social

Good decision and statement! 👏👏 "(...) preserving the integrity and reliability of the ACL proceedings is our responsibility and these measures are necessary to uphold the standards of our community and to ensure that the work published remains a credible resource for future scientific inquiry."

NEW: A serious staffing shortage — of the Trump admin's own making — is delaying the agency's ability to send billions to universities around the country, leaving labs reeling in the meantime. “I thought we were at rock bottom”, a senior NIH official said. “We are below rock bottom now.”

Two new facts stand out: 1. 85% of hallucinated citations in preprints are also in the subsequent journal version (thanks, peer review!) 2. Fake cites more likely to use the names of (male) scholars who are already highly cited, creating a fake-citation Matthew effect. arxiv.org/abs/2605.07723

UK 🇬🇧 report but lessons are probably applicable to universities in other countries.

An analysis of 141 films since 1919 finds geologists are usually portrayed as likable characters, though nearly a third of on-screen geologists die, most often by murder or geological events. doi.org/hb3tq4

Subskeet: pressgazette.co.uk/platforms/bl... www.emarketer.com/content/blue... www.softwareseni.com/why-bluesky-... www.similarweb.com/blog/insight...

Retractions and other publisher responses to problem papers doesn't affect citations. www.medrxiv.org/content/10.6... Quick thread. 1. There's no sentiment analysis. People might cite flawed papers to say, "This is a flawed paper." 1/4

WE’RE NOT DONE YET! 🎉​ ​ The PWHL is expanding to Hamilton, ON and Las Vegas, NV!​ ➡️ pwhlhamilton.com​ ➡�� pwhllasvegas.com�om​

Steeltown, we’re BACK - for good this time. 👏 📰 youtu.be/oEzj-wS4_6E

New, by me: U.S. lawmakers want answers from the chief executive of Instructure, which makes the Canvas software for schools, about the company's data breach and the defacement/extortion shitshow that followed. Millions of students' data stolen in the breach. Instructure paid the hackers' ransom.

Scientists have identified ancient proteins from Homo erectus, marking the first time genetic info has been extracted from the species. But the results raise the question -- Do we even know what Homo erectus is? 🧪🏺

Correction: while not listed on the main postdoc page, the Biology postdoc does exist…but only if you work with AI. bsky.app/profile/sedg...

On NSF’s postdoc page, the biology and earth science fields have been completely eliminated; the only remaining fellowships are in math, astronomy, engineering, and “entrepreneurship.” Weird choice not to support research on the life support system for 8 billion people. www.nsf.gov/funding/post...

FOR THE FIRST TIME EVER, THE VICTOIRE ARE HEADED TO THE WALTER CUP FINALS‼️

You know how many headlines use phrases like, “Person hit by bullet,” which makes it sound like an unfortunate accident with an inanimate object rather than the bullet being, you know, fired from a gun by police? Same phrasing here. Do not blame gen AI. Blame authors, editors, and reviewers.

I've long wondered about the design of coming up with drug names, and this episode does NOT disappoint! Drug naming is, as I expected, driven by a lot of considerations that you wouldn't expect - like how the name looks when it's printed.

This is the best idiom ever. 🦀 and everyone should be thinking about the immortality of the crab. (I must use this as a prompt to write a flash fiction) Source: en.wikipedia.org/wiki/Thinkin...

How are universities feeling about the agreement that Instructure reached with cyberhacker including a pinkie-swear not to use the data exfiltration for further extortion of universities, students, and faculty?

Dude *what*?!?

Using AI allows us more time, apparently. Yet we're busier than ever. There are more meetings that get recorded and AI summarises. Nobody watches the recordings or checks against the summary because nobody has time. It helps everyone look busy, but hides skills needed to transcribe, check, report

Slightly increase a band. ABCD.

Even vegetarians and vegans get enough protein without supplements!

Second-Hand Slopping. AI is like smoking. Even if you don't use it, you still have to deal with it every day.

Proud of my students for their research, which for the first time located the descendants of those enslaved by Columbia University's founders. www.columbiaspectator.com/news/2026/05...

Teen: Do I need to pay attention to this hantavirus thing? Me: Awww nah, you can skip this one. There will be another.

This result took 40 years to emerge. Not 3. Not 5. Not a funding cycle. Four decades. We’re trying to understand planetary systems with attention spans measured in quarters and election cycles.

The institution: University of Central Florida. The speaker: Gloria Caulfield, Vice President of Strategic Alliances, Tavistock Development Company. The audience: College of Arts and Humanities. Her reaction to being booed for loving AI: PRICELESS. Full video: www.youtube.com/live/zwYkHS8...

Companies are pitched that gen AI is like “a team of PhDs,” Not realizing how much of a bargain an actual team of PhDs is.

This article is an excellent example of the extent to which local news is now regime propaganda. All 4 people quoted are regime politicians giving regime talking points. Not one scientist, in a story about scientists.

Reminder that unis do not have to develop individual online learning systems to be safe! They just have to do sensible data sovereignty stuff like use the self-hosted open source version!! It’s completely feasible & we’ve been recommending it for YEARS!!! jilltxt.net/uib-self-hos...

"You're doing #SciComm wrong" says this article taking no-so-subtle digs at @standupforscience.net. Quick reactions: - Reminds me of the "framing" debate of the 2000s - Seems to assume that anti-science is organically rising bottom-up, not driven from top jcom.sissa.it/article/pubi...

Ram V told me the story of an assistant who watched Katsuhiro Otomo ink the black dome at the end of Akira by hand, crosshatching every millimetre. The assistant asked why. Otomo said, “There are a billion people dying under that dome. I have to count.” Software cannot fake that.

Frazer Irving named the missing thing better than any technologist; Ai output is too good. Too consistent. It doesn’t know where to be lazy. Real artists pour effort into faces and hands and let backgrounds dissolve. The machine wants to fill everything in. The artist knows what to leave out.

PKP Founder John Willinsky presented the Publication Facts Label (PFL) at the World Conference on Research Integrity this week. The PFL plugin brings article and journal publishing facts together in one place: pkp.sfu.ca/2026/03/31/p... #WCRI2026 #PFLabel #ResearchIntegrity #JournalIntegrity

The number of species named after David Attenborough really illustrates how important science communication is to science and scientists. Thanks to everyone doing the work to make the world a more knowledgeable place!

Told my kid I was having the kind of week where I just need to scream into the void for a while. He made me a void. You can use it too, if you'd like.

When your mutual with 50k followers asks for facts in your specific area of expertise late at night

Canvas, the Strait of Hormuz for American education

Alas, my favourite David Attenborough moment seems nowhere to be found online. It's in The Life of Mammals, where (if I remember right) Attenborough delivers a perfectly timed, "and the catch of the day... is salmon," and turns to gesture to a bear who obligingly catches a salmon RIGHT THEN.

This semester, I have been deeply frustrated with Moodle. But unlike Canvas, at least it hasn't been hacked...?

I’m aware this is news to CBS and this dude but we were literally shouting about this a year ago, it’s not unearthed, and it’s even stupider than you think: the cruise ships paid their salaries. It literally paid more than it cost. We gave money back rather than control disease because RFK

FDA blocked journal publication of COVID vaccine safety study in "act of censorship". But it's on medRxiv... www.medrxiv.org/content/10.1... www.nytimes.com/2026/05/05/u...

Zero-Click Readership: Are AI Overviews Changing the Way We Discover Research - The Scholarly Kitchen | Every AI summary of one of my papers has been misleading, missing the key point, or factually incorrect. Read actual papers, not garbage from the energy guzzling plagiarism machine.

“I think it would be fair to say it went poorly.” Saved you a click.

Who are your favorite science communicators who are academics (any field)? Who’s really getting public engagement right?

What would it take for Chinese clinicians to be judged on, I don't know, their clinical work and not be forced into trying to put their names on research papers?