HN user
I'm amazed ticalc.org is still alive and kicking. So much nostalgia. Joltima was what convinced me to learn assembly. So far ahead of its time on the TI-86. Full featured RPG with turn-based combat on a graphing calc. Glad the history is still accessible online.
Volta Labs (YC W19) uses this technology to prepare samples for DNA sequencing. https://news.ycombinator.com/item?id=40075571
While it’s not in the ARPA-H announcement, the broader press coverage is using the moonshot frame. https://www.nytimes.com/2024/08/13/us/politics/biden-researc...
Check out vtk.js (and VTK proper) as a fairly mature implementation of this. I haven’t kept up with the development but it was pretty far ahead when it came out years ago.
Yes. I don't use it heavily, but I have set up an account and added a moderate number of articles to it. It has an option to notify you of new, relevant publications that works pretty well. It tends to find articles my pubmed RSS searches / google alerts miss.
The second layer is hard. I tried something in this space in mid-2018. Full text extraction and sentence segmentation tech was adequate, but extracting the discourse tree and building the graph was a bit of a struggle (trying to repurpose a collection of academic/open tools to get something useful). Never published or released the code.
If interested, a few rabbit holes to explore (no affiliations):
https://scite.ai -> best option for citation mapping, but same issues you described above
https://www.semanticscholar.org and AI2 -> the best group working on tooling in this space
https://www.weave.bio -> early startup trying to build this out
The hardest challenge in my view is solving the intermediate representation issue. You have to establish a DSL/nomenclature that provides the range required to represent a complete scholastic discourse while also being computable.
As someone who just got the 7B running on a base MacBook M1/8GB, I strongly agree. The rate of tool development & prompt generation should see the same increase that Stable Diffusion did a few months (weeks?) ago.
And given how early the cpp port is, there is likely plenty of performance headroom with more m1/m2-specific optimization.
A biorxiv paper on the neural sewing technology: https://www.biorxiv.org/content/biorxiv/early/2019/03/14/578... - I believe the authors are now at Neuralink.
LateX Lab was/is exactly that: http://code.google.com/p/latex-lab/
A demo is up and running at http://docs.latexlab.org/
* not affiliated with the project in anyway, just remember using it awhile back.
^^ What he said.