Jay Patel

Preprinting + community review (an emerging practice) may be all that may be needed, so no need for journals in the traditional sense.

*just* a preprint server?

📅 Mark your calendars for #SIPS2027!
The 2027 SIPS conference, organized in collaboration with the Association for Interdisciplinary Meta-Research and Open Science @aimosinc.bsky.social, will be held in November at the University of Melbourne, Australia.

We are looking forward to seeing you there!

I would like this. And your eval criteria seem aligned with sleuths’ criteria. How hard would this be?

How popular is swimming in the Finnish winter? Big hobby?

Thanks for making and sharing! Happy to help assess its accuracy if you want to build something scalable. I saw a few papers over the past few semesters that had CI issues.

If only social media channels like Bluesky could do some kind of matching based in granular expertise. # ATProtoScience

Ah, yes. A perennial problem is how to source informal peer reviewers for such tasks. I'm happy to take a look, but not the most expert with that specific error. I wonder if Gelman might like to chime in. You could author an email to him.

Proceeding to do what? Post on PubPeer or author a comment in a journal?

In other news (plumbed from the depth of my LinkedIn reactions):

🙂‍↔️

A nice reform to support the publication or replications (a major issue):

1. plan and submit study to journal as replication
2. review
3. accept/reject based on the idea, not data
4. collect data
5. complete paper and submit for acceptance

Does this make doing replications more palatable to you?

Makes sense, the exams + holidays seems to constrain time. Will check out the recent data.

How is the trend by discipline? I've noticed much AI and CS discourse is on X, whereas Bluesky is good for everything else these days.

Also curious how geography (tougher to ID, I'm guessing) moderates the trend.

In December 2025, I did a quick analysis of Bluesky data and it seemed that research mentions/month were trending down. Are things changing now in January?

I'm curious to see the line graphs.

May they all be vetted by experts before being shipped.

Got a specific link? I'm curious what the reactions are.

Cartoon for @airmail.news

What made it a good paper, specifically?

👨🏼‍🎓 Me: Stews over the Theoretical Contributions section of the Discussion for months in quiet contemplation. Writing, rewriting, doubting, never fully sure.

🤖 ML researchers: "We offer no explanation as to why these architectures seem to work; we attribute their success... to divine benevolence."

Got the URL to his post about it? Or was it not posted publicly?

Bad or even a hoax article would be something

Congrats! Would it work on my M4?

Those Ivy League rates are something, huh?

Does it go to Vancouver from the East Coast?

Stealth corrections are not new news. RetractionWatch has reported this.

screenshot of preprint with highlighted Marx quote

I thought my #AcademicWriting was bold and spicy 🌶️

In this 🤖AI preprint, I find an example that asks me to level up my game:

“Karl Marx wrote that “the human essence is the ensemble of social relations”

Yeah, Zotero 8 has a special folder just for retractions native. Previously, I used a plugin. Good to see that kind of work. Now PubPeer and social media integration could be added.

Which initiative? Plugins or native features to detect retractions?

Very cool, will check it out.