M. S. AtKisson

Policy Notice: Implementation of Policy Changes to Proposal and Award Policies and Procedures Guide (PAPPG) 24-1, Supplement 1 This notice implements several revisions to NSF's Proposal and Award Policies and Procedures Guide (PAPPG) 24-1.

The source, and other changes: www.nsf.gov/policies/doc...

16. Research Misconduct Chapter XII.C is revised to incorporate the use of artificial intelligence (AI)-based tools. The underlined text has been added: RESEARCH MISCONDUCT means fabrication, falsification, or plagiarism, whether committed by an individual directly or through the use or assistance of other persons, entities, or tools, including artificial intelligence (AI)-based tools, in proposing or performing research funded by NSF, reviewing research proposals submitted to NSF, or in reporting research results funded by NSF.

NSF has updated the PAPPG, including this bit on misconduct. If your LLM use results in "fabrication, falsification, or plagiarism", it's on you.

Notice of Request for Information; Accelerating the American Scientific Enterprise The Office of Science and Technology Policy (OSTP) requests input from all interested parties on Federal policy updates that aim to accelerate the American scientific enterprise, enable groundbreaking...

Have you seen this RFI? Are your institutions responding? www.federalregister.gov/documents/20...

NIH has said the Personal Statement and the Contribution to Science will not go away with the shift to SciENcv. NSF moved the Synergistic Activities to a "biosketch supplement" page. This is why we suggest moving them to a separate document. No idea what the instructions will say on formatting.

More on changes: required Research Security Training. See the link in the page below for the condensed version. Certification of institutional completion is done by your grants office signing official (like "drug-free workplace" etc.)
grants.nih.gov/grants/guide...

before the trump admin, this would have been one of the biggest political scandals in history and alone would be solid grounds for impeachment and removal.

Yes! Me too!

Movie you’ve watched more than six times with a gif. Hard mode: no Stars (Wars nor Trek), LOTR, or Marvel, Disney Animated or Pixar.

Hmm..

IIRC, the trend of using modular budgets for NIH proposals has dropped to ~20% of applications. The drop in buying power is significant from the 90s to now

NIH says they will announce (finally) the updated biosketch requirements for proposal submitted on or after January 20, 2026 by December 15. Be prepared. Get your SciENcv and ORCID set up sooner rather than later.
grants.nih.gov/grants/guide...

NIH Biosketches? Time to re-tool yet again.
1) Go to SciENcv if you have not yet set up your profile and do it.
2) Go to ORCID if you have not yet set up your profile and do it.
3) Prep a version of your Personal Statement and Contribution to Science in a separate document.

For you NIH folks, two interesting announcements. 1) no more LOIs. 2) No more needing permission to apply with a budget of over $500K per year. grants.nih.gov/grants/guide...

this guy has long been my personal
this is why i wanted to do toxicology
type story

This is a quote from the article summing up both studies in Technology Review: "A multi-university team of researchers has found that chatting with a politically biased AI model was more effective than political advertisements at nudging both Democrats and Republicans to support presidential candidates of the opposing party. The chatbots swayed opinions by citing facts and evidence, but they were not always accurate—in fact, the researchers found, the most persuasive models said the most untrue things."

Two independent studies found that AI chatbots were better at persuading voters than political ads. The most persuasive bots also lied the most. This is something that humans working in psyops have known for decades. AI is psyops at scale. www.technologyreview.com/2025/12/04/1...

Clair Patterson - Wikipedia

Thirty years ago today, the scientist Clair Patterson passed away. He campaigned tirelessly and effectively to remove lead from gasoline, paint, pipes, and food cans. The drop in violent crime that correlates with this change? Massive.
Science hero.
en.wikipedia.org/wiki/Clair_P...

Happy repeal of Prohibition day!

This open letter against the uncritical adoption of AI in universities, spearheaded by @olivia.science and @irisvanrooij.bsky.social who have both published a whirlwind of excellent academic papers in 2025: openletter.earth/open-letter-...

a table with two columns. on the left are cells that say "reading a call is a non-linear process of creating and revising hypotheses about “what is wanted,” what is required, mapping that model against one’s network, and imagining proposal text. on the right are descriptions of behaviors and other elaborations corresponding to each cell.

day job hat on again

several years ago i did a little research project on how experts read Horizon Europe funding calls (mainly because I learning myself)

here's a key slide

When intelligence is defined as logic disconnected from human experience then all code is intelligent

When technology is assumed to be ontologically Good then any criticism of any element must mean rejecting the whole

I can see how these conflations arise

They're still fucking annoying though

"I think this president thinks we have the world's largest economy ever because of real estate investors. It's not—it's because of science."

Sen. Mark Kelly criticizes Trump's cuts to critical scientific research.

The cable, sent to all U.S. missions on December 2, orders U.S. consular officers to review resumes or LinkedIn profiles of H-1B applicants - and family members who would be traveling with them - to see if they have worked in areas that include activities such as misinformation, disinformation, content moderation, fact-checking, compliance and online safety, among others.

Holy shit.

Reuters reporting that new admin instructions on visas are if you worked at a platform in trust & safety or content moderation or on fact checking or online safety at an platform you *and your loved ones* are ineligible for H-1B visa.

www.reuters.com/world/us/tru...

This is also why (per a key plot point in the book), academia can be a disaster because we're frequently like everyone's dysfunctional family, and unless we train specifically to protect folks and overcome that dynamic, the default tends to be inward looking. Extremely bad.

Exclusive | Hegseth Asked Top Admiral to Resign After Months of Discord The relationship between the Pentagon chief and Adm. Alvin Holsey had been rocky since Trump’s inauguration.

Wow, Adm. Holsey didn't leave of his own accord, Hegseth pushed him out. What looked like the SOCOM head resigning rather than do war crimes now looks like SecDef canning him to facilitate war crimes.

Congress should have him testify, and would've already if war-crime-sympathetic GOP didn't run it.

This is literally the sales pitch from Gattaca

A short preprint describing our adoption of more consistent and precise use of terminology related to reproducibility, robustness, replicability, repeatability, and credibility.

osf.io/preprints/me...

What happens when ChatGPT cites to a real case but misstates what the case stands for? That doesn't fly for Judge Behm in Michigan. The court says that while a litigant could make the same error based on faulty reasoning, LLMs can't, because they can't reason.

And while a litigant might get away with similar overstatements because they could, perhaps, reason their way to showing how a case's stated holding might extend to novel situations, an LLM does not reason in the way a litigant must. To put it in a slightly different way, LLMs do not perform the metacognitive processes that are necessary to comply with Rule 11. When an LLM overstates a holding of a case, it is just piecing together a plausible-looking sentence

h/t @robertfreundlaw.bsky.social

holy shit, an accurate legal critique of LLMs. LLMs don't reason because they're just stitching together plausible-looking sentences indifferent to the content

What if instead of conceptualising disabled access in terms of special stuff to be doled out of a finite pot, we saw disabled access in terms of a shifted paradigm in which we just... made sure to build accessibility in from the start, for everyone, across everything, as much as possible

Adding alt text to this because this is fucking amazing and everyone deserves to read it.