Grilled Cheese

ExploreLog inSign up

Pat Schloss

@patschloss.bsky.social

24 Following2.3k Followers

Papa to 9 microbes, mackerel snapper, shepherd, master of sucking. I want to help you learn the skills you'll need to succeed in science.

PostsRepliesMedia
@patschloss.bsky.socialOct 15, 2026, 9:45 PM

I think this is also a problem with how we discuss AI. To me both were 100% AI generated, you just did more prompt engineering with what you're calling 100% Human. It seems that neither of these were a false positive then - rigth?

@patschloss.bsky.socialOct 15, 2026, 8:32 PM

I'd far rather read typos or honest grammatical problems than people presenting their skills as something other than they are. One reason that I went down this rabbit hole was that the English in these spammy emails has gotten 1000x better in recent years. I doubt that is the reality

@patschloss.bsky.socialOct 15, 2026, 8:29 PM

I'm confused - you said it was 100% written, but you gave the rules AI?

@patschloss.bsky.socialOct 15, 2026, 6:33 PM

What do you mean "some rules applied"?

@patschloss.bsky.socialOct 15, 2026, 3:24 PM

I have also followed up with the people who emailed me asking if they used AI to write the emails. Of those who have replied, they all acknowledged using something like ChatGPT to write/rewrite their emails to me

@patschloss.bsky.socialOct 15, 2026, 3:15 PM

Maybe I just write so poorly it doesn't flag my writing as AI 😂

@patschloss.bsky.socialOct 14, 2026, 3:36 PM

The latest from the Schloss Lab is a collaboration with @balunaslab.bsky.social. If you've been a fan of mothur and wished there something like it for MS/MS data, I present mums2! journals.asm.org/doi/10.1128/...

@patschloss.bsky.socialOct 13, 2026, 5:19 PM

Yeah. I just put through as many as I could do in one day on the free plan

@patschloss.bsky.socialOct 13, 2026, 1:05 PM

I don't really want to work with either category of applicant. Are any grad programs doing anything to guard against AI or seek disclosure from applicants? Is anyone modifying their process to open more interview slots?

@patschloss.bsky.socialOct 13, 2026, 1:05 PM

We get little actionable data from applicants. The personal statements give us a lens into how well they can communicate and think. Someone who gets flagged as using AI is telling me they cannot communicate. If they won't self-disclose (and are truly using AI) then they tell me they can't be trusted

@patschloss.bsky.socialOct 13, 2026, 1:05 PM

I ran 8 of the most recent emails I've received from people wanting to work in my lab through pangram. 1 came out as written by a human, 2 mostly AI/some human, and 5 100% AI. I'd love to hire that 1 person just out of principle. Are any grad programs running application materials through pangram?

Photo, no description
@patschloss.bsky.socialOct 13, 2026, 12:18 PM

I feel like we've already solved a big chunk of lung cancer with actual intelligence... don't smoke. 🤣 What am I missing?

@patschloss.bsky.socialOct 9, 2026, 1:23 PM

Belt and braces to hold up his pants, but only winged tips to hold on to the beam? 🤣

@patschloss.bsky.socialOct 2, 2026, 7:27 PM

I'm in awe that you all still have chalkboards and haven't been forced to use dry-erase boards. What I wouldn't give for a nice chalkboard

@patschloss.bsky.socialSep 30, 2026, 7:13 PM

This is my existential fear of asking my kids anything discipline-wise and knowing that they would be correct

@patschloss.bsky.socialSep 30, 2026, 4:33 PM

Returning to the paper that started this series of papers, I wonder if 3 papers in response to the reviewers will be enough 😂 By the way, the preprint of that paper has now been cited 26 times! www.biorxiv.org/content/10.1...

@patschloss.bsky.socialSep 30, 2026, 4:33 PM

In 2023, I dropped this in mSphere showing that the original anti-rarefaction harangue was poorly done (w/ no responses from original authors) journals.asm.org/doi/10.1128/...

@patschloss.bsky.socialSep 30, 2026, 4:33 PM

In 2024, I dropped this in mSphere journals.asm.org/doi/10.1128/...

@patschloss.bsky.socialSep 30, 2026, 4:33 PM

Today I dropped this in mSystems (w/ no responses from RPCA authors): journals.asm.org/doi/10.1128/...

@patschloss.bsky.socialSep 30, 2026, 4:33 PM

In response to these comments, I have now published *3* papers saying that rarefaction is the way to go...

Older posts
Terms of UsePrivacy PolicyCommunity StandardsHelpGet the app

Grilled Cheese is a product of Village Compute

Version devBuilt at: 2026-10-16 08:00:47 EDT