Critical mass of high ranking people in USGOVT are all saying the same thing: UFOs are real and we've been studying them in secret for decades. The scientific community needs to engage and get up to speed youtu.be/DkU7ZqbADRs

It’s finally here!! See everyone tomorrow!

27 days until the spatial summit! We’ve opened up a new block of spots if you weren’t able to register before. Get them while they last!

Registration for the 3rd Annual Spatial Biology Summit is now open! Don't miss out! Hope to see you in September.

A recent podcast featuring @matthew_pines does a fantastic job of explaining from a national security and intelligence perspective why these moves by Congress are so unprecedented and shouldn’t be discounted: youtu.be/DJJM4YydWkI

Tried out @Meta Segment Anything on tissue images, really impressive. Automatically identify multicellular structures-i.e. glands, ducts, arteries, etc. Segmentation below is with no user input. Not perfect but pretty close, definite time saver over manual annotation.

Arun - it’s a chest x-ray generated by #DALLE

To all the radiologists, this image was shown at the hubmap meeting for a patient with shortness of breath. What is the diagnosis?

@NoahGreenwald @CandaceCLiu check it out! Soufflé part 2! TOOT TOOT!

Just tried @Meta’s new AI chatbot #blenderbot. It keeps telling me how it’s friends think it’s funny, but it won’t tell me a joke, I feel like it’s trolling me.

This interview is important. You should watch it. Proud to have been trained by @GarryPNolan, who is dragging the topic of #UAP into the mainstream. youtube.com/watch?v=u7cKhIJn…

Thank you to @BioRender for making this wonderful cover art for our paper cell.com/cell/fulltext/S0092…

I wish I were smart and funny enough to have thought of this

Extremely excited to share our work mapping the human maternal fetal interface in the first 20 weeks of pregnancy. Thank you to @_hubmap, @stanfordPath, @stanfordMed, congratulations to @ ShirleyGreenba1, @ErinSoon_, @Inna_Averbukh on a beautiful high dimensional imaging study.

We found that the pathologist evaluations were all clustered around 0, indicating that there was no difference in preference between our predictions and those of the annotators. This represents the first time a tissue segmentation algorithm has reached human performance. 4/5

In fact, this accuracy was what motivated us to perform what I think is one of the coolest analyses of the paper: a head-to-head evaluation by certified pathologists, where we blinded them to the origin of different segmentations and asked them to pick the more accurate one. 3/5

One of the most exciting things about Mesmer, our new algorithm, is that it allows us to accurately capture tissue morphology. When we applied it across a range of tissue types, it was easy to see that it was accurately capturing the underlying makeup of these distinct cells. 2/5

Day 1 of #lockdown: already see where this is going. get board--> bake-->can't give baked good to lab-->eat baked good-->Repeat Case in point: Tried to make a soufflé last night, burnt, still ate it. Current models predict exponential growth.