This reminds me a bit of Wei-Hwa Huang's blogpost about secondary goals in board games - onigame.livejournal.com/34424.html I wonder how much this behavior is boredom / frustration / abandonment of the "intended" goal of learning as an LMS dictates...

I imagine, it will require more sophisticated prompting to get the AI version to do something unethical.

“Advice that hides tradeoffs shields people from making informed choices.”

Wait, I have a better summary: "subjects who practiced in context A had better test performance in context A than subjects who practiced in context B and then were switched into context A"

To me, "trust in science" is too large and too complex of a concept; and so is "scientists engaging in politics". It seems obvious to me that one can drum up evidence for effects in any desired direction depending on how one operationalizes the concepts.

Further thoughts: if the variable is one where changes should be interpreted proportionally (+/- 10%) then, of course the y-axis should include 0.

I think my latest pet peeve is folks convinced that the y-axis in plots ALWAYS needs to start at 0. It is necessary when the plot comparison is LENGTH. (e.g. the height of bars in a bar chart) It is NOT necessary when the plot comparison is POSITION. (e.g. points or a line chart)

I'm not and never saying AI is good. But the retreat back into romantic genius ideology that it's prompting, I will say, is bad.

Question: Is there an existing way to create a schema for a CSV file that a machine can use to understand and/or validate the data?

🚨🚨🚨 Six universities will receive $250K each through the Modernizing Academic Appointment & Advancement (MA3) Challenge to pilot bold reforms in faculty hiring, evaluation, promotion, and tenure systems. Today @orca-open.bsky.social is excited to announce the inaugural MA3 awardees. [1/5]

" @ucsantacruz.bsky.social will develop and implement institution-wide guidelines to evaluate open science, open source software, science communication, and graduate mentoring in merit and promotion." 🙌🏽🙌🏽 proud to see our VPAA, Christina Ravelo, spearheading this effort!

I made a whole workshop based on this paper uf-repro.github.io/data-organiz...

"Because nobody is required to change anything. Because the indifference is structural. Because fuck you, but politely, with a ticket number." fireborn.mataroa.blog/blog/because...

100% I don't think any previous technology or experience has really prepared humans to deal with "errors" in LLM output. (error is in quotes because the notion that an LLM can produce an error feels rather like suggesting that a magic 8-ball can lie to you)

I am not sure quite how it happened, but after the cameo in Into the Spider-Verse, an eight-episode live action series of Nic Cage as Spider-Noir was greenlit and made and is coming out in May??

Key to efficient learning is realizing how we ACTUALLY learn, not just what FEELS like learning. I wrote a Claude Skill for some friends to help them think about this and they've liked it -- see Principles for some directions you could explore github.com/DrCatHicks/l...

"move fast and break things" has never respected the maintenance and glue work required to keep systems running.

Does anyone else get tripped up by formatting links in markdown as [text](url)? I often first write it the other way around as [url](text)... maybe I've had too much formative time with writing html links as <a href="url">text</a>

Constructively, a lot of the replies seem to feel seen (the pain?) of this post. It can be hard to write up Bayesian credible regions when you know your paper is headed for reviewers who will write back "yeah but is it significant?" Here's my brief advice:

If you, a 🧪, know a scientist involved in this, sit them down and explain the Nuremberg trials and the Doctors’ trial en.wikipedia.org/wiki/Doctors...

I didn't do real fieldwork until I got into my postdoc at UF. But it felt important to: a) develop respect for the work that goes into collecting data b) develop respect for the data and the unwritten uncertainties around everything related to counts, IDs, measurements, etc.

Depends on what you mean by "wrong". But I'd say that probably close to 100% of papers are wrong in some form or other.

movie pitch: parent trap reboot, but it's bringing together Patrick Rothfuss and Brandon Sanderson to finish The Kingkiller Chronicle

this, from @kevinbaker.bsky.social, is a better analysis of the intersection between LLMs and academic science than 98% of what's out there.

A lot of people (professional scholars included) start a literature search with a claim that they want backed up by a citation. LLMs are designed to give answers that please users, so surprise surprise, they often make up citations when that's what you ask of them.

The question about when abstract writing should occur is really interesting! It illuminates the different objectives of an abstract, and what is achieved through the writing process. /1

My partner and I have also been exploring spotify alternatives; also settling on Qobuz (for streaming). (and always a +1 to bandcamp if you like owning copies of specific albums)

KPOP Demon Hunters is a movie about the importance of effective public health communication in this essay I will...

Kudos to the scriptwriter of the faculty meeting scene in S2E3 of A Man on the Inside. The way everyone got distracted from the agenda item to ask questions about a blazer is spot-on!

1/ Egalitarianism should begin at home. I link to this article by @bencasselman.bsky.social in light of the communications between Larry Summers and Jeffrey Epstein that have just been released. The released emails and the fact of friendship are vile. www.nytimes.com/2021/02/23/b...

“What these pundits are nudging us to do, ever so politely, is accept that women, in the main, are accustomed to being a little degraded, a little underpaid and ignored and dampened in their ambitions, in ways that men are not and never will be…What a gift it is…to have no choice in the matter.” 🔥🔥🔥

I think this take on the sandwich guy trial is correct. our system is built assuming that prosecutors will show appropriate restraint. the sandwich guy case went on for way too long and that in and if itself should be raising alarm bells www.lawfaremedia.org/article/sand...

I think jobs should always list salary and that people should talk about their salaries openly. Your salary is not your worth. It's just what you get paid. This would help unite workers and reveal any nefarious pay gaps.

Your reminder that there are plenty of folks involved in research integrity that don't write racist / xenophobic click-bait.

much is made of mamdani's digital strategy, but all of that Posting went to proving that he lived in the real world, appealing to real people about real issues. this dynamic is something to watch as GOP (and center libs tbh) drift into the Posting ether of the X filter bubble

things like Jerry Falwell was hoping to end public education in the US because it is integrated really situate in context the current destruction of the department of education

3/3 Theory&Society, you'll recall, was taken over ~2 years ago by a group closely associated with 'heterodox' scholarship. The piece fits my continuing impression of much of the heterodox crowd that they're very much willing to lower standards if it fits _their_ ideological predispositions.

YES YOU SHOULD VOTE IN LOCAL ELECTIONS TOMORROW EVEN IF MAMDANI IS NOT ON THE BALLOT WHERE YOU LIVE

“Multiverse analyses require thorough, theory-based model selection. Otherwise, they become a ‘dangerous tool’ that drowns valid models in misspecified ones, needlessly eroding trust in science.” A plea for thoughtful models by @kauspurg.bsky.social #MetaSci

I’m in a room where a whole ass faculty member said “Chat GPT 5 would be able to write a better NSF proposal than me” and… well… My inside thoughts got the best of me BC I said loud as fuck “that’s unfortunate, because there’s not a machine on earth that writes better than me.”

Due to this policy, I have just applied for a visa waiver to a country of which I am a citizen. Why? Because my American passport has an X on it, and now I can only safely enter the country using my Chilean passport with a binary sex marker. www.visaverge.com/passport/cbp...

I actually think the inverse problem (positive bias for positive results) is more of an issue? With a null result, (ideal) peer reviewers are understandably thorough in checking the methodology to rule out errors that would have confounded the results. /1

In the Trump regime, every accusation is a confession. The terrorism is coming from inside the White House. www.theverge.com/policy/79051...

This is TNC, and this is the foundational difference between his politics and Kleins’s. At a basic level you either believe in the biologically deterministic myths that historically and currently animate hate and exclusion and oppression, or you fight them with everything you have.

This fall, @megstampede.bsky.social & I are collaborating to create a 2026 calendar! Each month features a different animal, facts about them, & actions we can take to help them! Sales will fund our '26 Native 🌱 Project Order before 10/1 to get a free plant sticker! www.etsy.com/listing/4366...

IT'S OFFICIAL: Connecticut, Maine, Massachusetts, New Jersey, New York State, Pennsylvania, Rhode Island and New York City have formed the Northeast Public Health Collaborative. The collaborative will issue their own vaccine recommendations and coordinate public health efforts.

I’m sure she totally means Palestinians and their comrades speaking out against Zionism

#datalibs #medlibs NIH Data Management and Sharing Policy question - is data that was produced from grants that are now terminated still subject to the original grant terms and conditions?