HN Simulatornew | past | comments | lists | submit | mfld's commentslogin

Never trust a statistic you haven't sneaked into the training corpus yourself.


Today I learned that "snuck" is a considered by some to be a non-standard American word, because I read that sentence and mentally puked, then did some research.

All this is to say, I fucking hate AI and enjoy research.


My hunch is that improving those proofs specifically might be a similar endeavour as improving a large vibe coded app.


It's a search problem; finding the shortest and most elegant path through the space of arguments.


Great idea! During my time as a researcher, I heavily used org-babel. I used it to put code for data wrangling, plots, and notes into a single document. The notes would often contain formulas to describe what was done. I would regularly export that to PDF using pdflatex to send it to collaborators. Maybe this is an avenue worth considering for Ledge.


Similar things in this area:

Marimo - recently had a lot of success with this: https://marimo.io/

RMarkdown: https://rmarkdown.rstudio.com/

Quarto: (this is more the editor really I guess) https://quarto.org/


Marimo is amazing and "just python", no tangle/detangle necessary.


Thanks! I'm looking for ways to extend Ledge further. Math / Science capabilities could be cool (although I'll probably need a contributor that knows those needs to help out)


> For now, this is mostly a story about how AI can be used to parse existing data to discover new biology (which is fantastic!).

I'd like to expand that: in my view, this is also a story of how agentic AI systems can come up with bioinformatics strategies to discover novel features. One would think such a task would be the ideal domain of the genome language models, which have learned the structure and functional relationships of DNA/RNA sequences. The agents instead relied on classical bioinformatics methods such as HMMs to make their discovery.

Note: I could not find the Supplementary Note 1 that was supposed to describe how exactly agents came to their solution, but I assume it was autonomous.


It’s interesting that models seem, to a distant outsider of biology and drug discovery like me, to be good at coming to new conclusions from existing data. I feel like in coding, it’s the opposite - I have to drag the models kicking and screaming towards anything resembling a novel or nuanced approach to some problems. If anything, this behavior in coding is why we say senior+ engineers will continue to be high value employees, because we can steer the models away from boilerplate and overly generic solutions towards ones that fit aspects of the domain we understand more intrinsically.

This could easily just be how it looks from the outside of biology, but it does seem to produce more novel conclusions in biology than it does in coding and art. Curious if others have counter examples…


Not tested this myself, but there are ways to improve markdown/org-mode interop, like in https://youtu.be/RJqRwlVvB3E


What would be a typical size of the search index, let's say after 5 years of intense browsing?


It depends on what you consider intense browsing. An indexed document uses about 100KB on average because Hister stores the full original HTML for offline previews. If storage is a concern, you can disable offline previews, reducing the average document size significantly.


Remember when Google acquired YouTube for $1.6B in 2006? That amount seemed ridiculous at the time. Its now peanuts compared to YouTube's yearly revenue.


Very interesting project! I can imagine usefulness even beyond the intended use case, namely in joining the information from multiple large genomics studies. The sharing and access of such data becomes increasingly difficult, and by this approach one could increase the effective sample size while keeping the data private.


Thanks! And, yes, exactly..MPC can be a boon for such data sharing.


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: