Jan Domanski

A Cuban mechanic, Juan Carlos Pino, modified his 1980 Polski Fiat to run on charcoal instead of gasoline, offering a cheaper alternative amid fuel shortages caused by US sanctions reut.rs/3NzcSyP

"NOT MY CIRCUS NOT MY MONKEYS" IS ORIGINALLY A POLISH PHRASE??????????????????????

Excellent post by Rachel Thomas in the ways AI hype is outrunning reality in life sciences linkedin.com/feed/update/ur…

I had no idea this existed wow super handy

@nlarusstone Just compute it bro c’mon ;)

Tech confidently commenting on biology (and being wrong) remains unbeaten

@anshulkundaje @garrytan These seem like very odd bets. I think they have an ideal profile/phenotype of a founder and just rinse and repeat that. More broadly, (in my experience) VC are only moderately interested in the science at that early stage. Ability to sell to investors/customers matters most

Jacob is being very kind here. @garrytan (who has blocked me) & YC seem to be astonishingly uncalibrated & unaware to be funding startups like these. It's just embarrassing at this point.

Sorry to be the downer because this is an impressive story in some senses. But it is ~trivially easy to make a single mRNA vaccine. It's not hard. I cure mice of various cancers with various therapeutics all the time. I've made mice lose more weight in a month than tirzepatide does in a year. What is hard and expensive is proving its BOTH safe AND effective **in a randomized and controlled study in humans** while ALSO manufacturing it at clinical scale and grade. I am happy for this man and his dog. It is impressive. But y'all are overhyping it.

Whoa claude can now do visual explainers... hot-daaamn

New OpenFold3 preview out! (OF3p2) It closes the gap to AlphaFold3 for most modalities. Most critically, we're releasing everything, including training sets & configs, making OF3p2 the only current AF3-based model that is functionally trainable & reproducible from scratch🧵1/9

@DdelAlamo Protein ligand is absolutely data starved too, this is sus and I’ll have to think about it now thanks :P

Protenix trained an identical model with way more training data (2025 cutoff instead of 2021), demonstrating that antibody-antigen modeling, but not protein-ligand modeling, is currently data-limited (DQ SR % means % DockQ≥0.23) with this architecture

the IKEA effect is just strong with vibe coded things :D

my living costs of living in poland €1200 - vitamin d supply €1400 - antidepressants €1200 - therapy €990 - electricity for running 10k lux lights everywhere to simulate sunlight) €3900 - ai credits €2400 - printer filament help me budget this my family is starving

@tropianhs 🤌 ottimo

True story happened to me not long ago Senior Data Engineer interview with HR Freelance contract "What's your rate?" "400/day" "Oh no too expensive..." "What's your budget?" "150/day" "..." "Ok we tried" Guess the country

guys you're never gonna believe this

@chrishayduk @karpathy Love this, @open_fold this project is very cool

Currently cooking up minAlphaFold2, inspired by @karpathy's minGPT. We need clearer educational tools to help top talent learn bio AI and push the field forward. Will be starting with the Nobel Prize-winning AlphaFold2. AlphaFold3 coming soon. github.com/ChrisHayduk/mi…

@KamilTamiola @aaronmring @hkanji Spicy 🌶️

@aaronmring @hkanji @jandom_random hear hear!

Isomorphic must be in a crazy bind right now. Given the hype about their models/frontier capabilities, I can't see how they could reveal their actual pipeline without it being a huge let down. Heading towards: "solved biomolecular design and all I got was a me-too TL1A mAb"

whyyyyy, why can't we have nice things 😭😭😭

@intern Same for BD team watching engineers use a spreadsheet

engineers watching the BD team use claude code

Cutting edge AI use takes different forms in my day

@MichaelAArouet Being a nice place to live. Undervalued for sure.