When the panning starts, there is a difference in FPS between the cilia and the lower part

The prompt is the raw data

So the green cillia are "legit" images, while the rest is hallucinated, which makes sense given that cillia are oversized compared to the cells but somewhat of consistent size relatively to the scale bar they added afterwards (not a cillia expert, correct me if I'm wrong).

Some E. coli for scale... 🧐

...and by safe, I mean that it becomes too big to cross the outer membrane of E. coli/Salmonella, which makes it pass the Ames test as non-mutagenic. Yay !

It is the most dangerous molecule EVER. But don't worry, it becomes magically safe when you bridge two such molecules with a linker (eg GelRed)

Et demander d'avoir un partenaire <thèse+10, c'est bien beau mais on ne sait pas si et combien le CNRS et autres EPST en recruteront dans ces prochaines années qui s'annoncent rudes.

2026 : on rend tout le monde inélligible aux demandes de financements
2027 : on se rend compte avec stupeur que le nombre de demandes a chuté... donc on définance encore plus !
On va vers un démentèlement pur et simple de la recherche.

wait, so that gel was initially 0.35 cm and is now 3.5 METERS ??? 😱

Two-part epoxy could be a good alternative, it should cure clearer than CA glue and with a refractive index close to that of glass.

And a complete heresy to mix it in cocktails !

www.addgene.org/browse/artic... (Available from Addgene @addgene.bsky.social )