Stop glamorizing the hustle and start glamorizing the mussel

I know feminist linguistics is not super popular here on x dot com these days which is why I am very brave to say that “germinal” is objectively just a great word and should be preferred to “seminal” when describing something original and influential

There are now thousands (at least) of law-abiding biologists who distrust commercial LLMs Building off the grid is normalized. Bad actors who avoid security checks will go unnoticed by their peers Over-refusal creates a biosecurity risk by eroding social trust

Introducing EpiBench, an agentic benchmark for practical epigenomics analysis. 106 evaluations span CUT&Tag/CUT&RUN, ATAC-seq, ChIP-seq, and DNA methylation workflows. The best agent–harness pair passes 45.0% of evaluations. Evaluations reflect the assay outputs scientists use in practice. A task may depend on alignment files, peak calls, methylation tables, QC metrics, sample metadata, genomic annotations, or downstream summaries. Solving them requires a mix of coding, data analysis, and scientific judgment. Ground truth is hard to define even for short-horizon scientific tasks. Alternative task interpretations can produce multiple plausible answers. Candidate tasks are hardened through manual quality control. We remove prompts that over-specify the method, answers that can be solved with general literature knowledge, and ground truths that fail to reproduce under peer reproduction. Short-horizon tasks are the current frontier for scientific agents in epigenomics. Before models can own deeper biological reasoning, they need to become reliable at local assay-specific decisions.

The writing's on the wall

Applies equally to tweets, Jake

RTing this to prove that I've been bitching about bioAI refusals since before it was cool

I choose to believe this is satire

I think one reason why microgrants are disproportionately effective is that they act like a kind of market signal It's easy to say "that's interesting" "That's interesting," plus even a tiny bit of money, is a much better sign that what you are doing is actually interesting

"I'm trying to free your mind Neo. But I can only show you the door. You're the one that has to walk through it"

I beat it, three emoji and Fable flags it

I did my PhD in Harvard’s systems biology department - seeing great mathematicians and physicists routinely getting owned by the challenges of doing biological experiments shaped my views on uplift. Many of our levers for biosafety are in atoms not bits

If you get the words “powerhouse of the cell” tattooed on your face then AI facial recognition will refuse to work on you

New malware takes advantage of LLM biosafety refusals to disrupt AI-assisted virus scanning. What a time to be alive! socket.dev/blog/mini-shai-hu…

I've never felt more motivated to make sure open source wins across the board on every f'in thing. As the younglings say, LFG!!

I got curious about the real story of Ozempic Gila monster spit people cite when advocating for basic research funding. The truth is more interesting, and shows us more about the stories we tell ourselves about science than it convinces people to maintain the funding status quo

Especially important for Massachusetts to get biotech added to the COINS Act list of strategic technologies by supporting the bipartisan Biotech Investment National Security Act (BINSA) act proposed by @RepDebDingell and @RepMoolenaar ! Great opportunity for our Congresspeople and Senators to keep jobs onshore in one of MA’s biggest industries ! @JakeAuch @sethmoulton @SenWarren @SenMarkey @AyannaPressley @RepMcGovern @WhipKClark @RepLoriTrahan @RepRichardNeal @USRepKeating @RepStephenLynch Op-Ed in Globe linked below 👇

Can't believe we got pig semen eye drops before GTA 6

Cretaceous noclip

I'd argue that these observations are all spot on, but that bio databases are hard for agents to traverse in part because they are *also* hard for bioinformaticists to traverse. Treating data as secondary to supplemental to publications got us here - empowering bioinformaticists with agents will help us progress faster