Lightnews — Scholar-powered news

Will Ratcliff

@wcratcliff.bsky.social

6.8K followers 3.9K following 1.4K posts

Evolutionary biologist (Multicellularity & social evolution). Prof. at Georgia Tech & Director of the QBioS PhD program. https://ratclifflab.biosci.gatech.edu/

ratclifflab.biosci.gatech.edu

Posts Media Videos Starter Packs

Pinned

Will Ratcliff @wcratcliff.bsky.social · Mar 5

1/46 Hey folks, we have a new paper out on the MuLTEE. Strap in and I’ll tell you the story of how this “little paper on polyploidy” turned into the most data rich paper our lab has produced, largely thanks to the leadership and work ethic of @kaitong25.bsky.social.

www.nature.com/articles/s41...

Genome duplication in a long-term multicellularity evolution experiment - Nature

In the Multicellularity Long Term Evolution Experiment, diploid yeast evolve to be tetraploid under selection for larger multicellular size, revealing how whole-genome duplication can arise due to its...

www.nature.com

16 170 360

Reposted by Will Ratcliff

Ricard Solé @ricardsole.bsky.social · 6d

Looking forward to speak at this @embo.org Workshop event meetings.embo.org/event/25-mul... to explore the origins of multicellular complexity along with great colleagues @multicellgenome.bsky.social @wcratcliff.bsky.social @viktri08.bsky.social I will be discussing synthetic multicellularity.

4 23

Will Ratcliff @wcratcliff.bsky.social · 10d

Who’s trying to poison T. H. Huxley?

Matt Novak @paleofuture.bsky.social · 10d

Darwin’s dog food has tested positive for salmonella and listeria but is reportedly being uncooperative with the FDA.

“Arrow Reliance has repeatedly failed to issue recalls of its products under similar circumstances in the past.”

Darwin's pet food tests positive for Listeria and Salmonella

The FDA has again found bacterial contamination in Darwin’s Natural Pet Products dog food and the producing company, Arrow Reliance, is again refusing to

www.foodsafetynews.com

1 16

Will Ratcliff @wcratcliff.bsky.social · 11d

Will do!

Reposted by Will Ratcliff

Roland Hatzenpichler @environmicrobio.bsky.social · 12d

Our article on equitable reuse of public sequencing data is out! Led by @alexjprobst.bsky.social, @lhug.bsky.social, Cristina Moraru, @geomicrosoares.bsky.social, @folker.bsky.social and myself -, co-authored by Anke Heyder, and developed in consultation with 167 scientists. tinyurl.com/n6yeanmk

A roadmap for equitable reuse of public microbiome data - Nature Microbiology

In this Consensus Statement, a consortium of microbiome scientists discuss current sequencing data sharing policies and propose the use of a Data Reuse Information (DRI) tag to promote equitable and collaborative data sharing.

tinyurl.com

5 22 39

Will Ratcliff @wcratcliff.bsky.social · 11d

Check out what I got in the mail- the latest from @joanstrassmann.bsky.social! Perfect timing, I’m about to hop on a transatlantic flight.

2 1 17

Will Ratcliff @wcratcliff.bsky.social · 13d

Commentary coming asap!

Will Ratcliff @wcratcliff.bsky.social · 14d

I did it back in 2015, but now can't find the time lapse video!

Will Ratcliff @wcratcliff.bsky.social · 14d

New lab photo just dropped

1 36

Reposted by Will Ratcliff

Maik Bischoff @maikbischoff.bsky.social · 14d

1/12 A very special moment for me! 🎉

My first paper as corresponding author—a @jcb.org Perspective on how contact-based decision-making in collective cell migration can itself encode blueprints for complex patterns and shapes.

rupress.org/jcb/article/...

#cellbio #devbio #science

🧵for details

Cartoon of a speculative heterogenous cell population performing mesenchymal collective cell migration.

4 29 100

Will Ratcliff @wcratcliff.bsky.social · 18d

Awesome! Thanks for the info!

Will Ratcliff @wcratcliff.bsky.social · 18d

Is this a butterfly murder scene? Chrysalis above the red drops (likely gulf fritillary).

1 6

Will Ratcliff @wcratcliff.bsky.social · 19d

Writing is programming for human brains. As a PhD advisor or referee, I’m the compiler, and it’s my job to tell you where it threw an error (and how you should adjust it so it runs!).

3 35

Will Ratcliff @wcratcliff.bsky.social · 20d

and editor!

Will Ratcliff @wcratcliff.bsky.social · 20d

Could be- I have not seen that happen however.

I think LLMs will turn out to be neutral to mildly beneficial for academic writing. Most people learn scientific writing by writing up their papers, and I expect people will care enough to work/think super hard for the foreseeable future.

Will Ratcliff @wcratcliff.bsky.social · 20d

At least early on, the quality of your writing is proportional the pain you experience writing it! If it's not painful, you're not thinking hard enough!

Will Ratcliff @wcratcliff.bsky.social · 20d

I find this reassuring. LLMs don't replace the need to learn how to write whatsoever.

There really are no shortcuts to thinking super hard for many hours. That is the only wait to do good science. All shortcuts are an illusion.

--> This is good for the humans! Roll your sleeves up and think.

3 1 11

Will Ratcliff @wcratcliff.bsky.social · 20d

I have seen some students with relatively poor English skills develop absolutely top-quality papers with the help of LLMs, but in every case, they are brilliant people with a very well developed idea of their work/paper arc.

I've seen even more students whose work was not improved at all.

1 6

Will Ratcliff @wcratcliff.bsky.social · 20d

The interesting thing with watching students learn to write, in the 'genie out of the LLM-bottle' era, is that avg. writing isn't improving as much as I would have expected.

To me this is evidence that knowing how to write well boils down to having a very clear idea of what you want to say.

2 4 23

Will Ratcliff @wcratcliff.bsky.social · 21d

“I fear not the man who has transferred 10,000 cultures one time, I fear the man who has transferred one culture 10,000 times” — @relenski.bsky.social, probably

3 7 73

Will Ratcliff @wcratcliff.bsky.social · 22d

Thanks for the invitation! I had a lovely time- outstanding questions too!

lflarrondo.bsky.social @lflarrondo.bsky.social · 22d

What a treat !! Last week we had the wonderful opportunity to enjoy an e-talk by @wcratcliff.bsky.social at our Faculty of Biological Sciences at the Universidad Catolica de Chile! Thanks a lot Will !

1 7

Will Ratcliff @wcratcliff.bsky.social · 22d

bsky.app/profile/wcra...

Will Ratcliff @wcratcliff.bsky.social · 22d

Oh, right- did all this with OpenAI's codex on GPT-5-high, using the standard $20/month plan.

I don't want to share this script, it's using my specific version of Conda & dependencies, and I have not yet sorted through my thoughts on the ethics of it.

Will Ratcliff @wcratcliff.bsky.social · 22d

It only assesses the transcript, so it should be voice voice agnostic. The funniest thing I heard was the AI chop a part where the hosts discussed an ad!

Will Ratcliff @wcratcliff.bsky.social · 22d

I forgot to say: I started making this while waiting for a dentist appointment today. I listen to podcasts to distract me, and was bemoaning the fact I had to listen to the ads.

I was like “huh, wonder how hard it is to cut them all out?”

Will Ratcliff @wcratcliff.bsky.social · 22d

But still..it feels like a sea change. Personal (shitty) software is accessible in a way it just wasn't before, which is quite exciting.

2 2

Will Ratcliff @wcratcliff.bsky.social · 22d

1 4