feliixh

I just hacked together https://www.firstbranch.ai in a few days, conversational intelligence for congress hearings.

I was talking about healthcare price transparency with a friend. He pointed me to a hearing in congress and I was surprised by how much I learned from one of the witnesses' testimonies so I built a "serendipity engine" for congress hearings.

I have been working on https://www.accessmrf.com - a catalog of all American health care prices. Since there's way too much data, I am building a taxonomy and deduplicating it extensively first, with the idea that I can shrink it by ~4 orders of magnitude and make the problem tractable. I have written a bit about how I do it here: https://www.felixhaba.com/writing/simplifying-healthcare-pri...

My goal this week has been to index all insurers that publish machine readable files. For now, my project helps researchers and founders in the space access the data via API, it is not consumer-grade yet. But if you are interested in this data I would love to brainstorm ideas or share my knowledge. Shoot me a message.

I am working on https://www.accessmrf.com a catalog for all health care prices published by insurers under the Transparency in Coverage rule.

I recently wrote a blog post using Min Hashing to estimate that at least 90% of the 1.17 petabyte dataset are duplicates and I keep investigating new ways of making this dataset manageable. https://www.felixhaba.com/writing/simplifying-healthcare-pri...

I've been building out https://www.accessmrf.com/sources to index Transparency in Coverage files across insurers in a single place. Most recently I have been looking at the overlap across files using Min Hashing and found that there is a lot of duplicate data. Currently, I'm figuring out a way of making this dataset more digestible by identifying and eliminating the 90+% duplicate and ghost data.

2,3,8,9

I'm building a catalog for health care price transparency data that aggregates the rates published by all insurers, to put everything in one place and make it easier for developers / researchers to access this data. https://www.accessmrf.com/

Very useful!

I'm building a catalog for health care price transparency data that aggregates the rates published by all insurers, to put everything in one place and make it easier for developers / researchers to access this data. https://www.accessmrf.com/

Great job, I intend to reproduce this on a similar dataset I've been collecting!

I will say, it would be great to see the color labeling done on domain url alone, to see how much of the topography of the map is driven simply by the different formatting characteristics of the websites you're gathering data from.

Def a psyop. Classic Catalonian move of seizing independence by writing self determination laws in code and carefully introducing a bug that they can then exploit to secede.

Haha, fair point, what I really meant is that LLMs will translate natural language to code, so building will be mostly in English while debugging will still happen in code.

One thing I think will dominate in the future is to write software documentation geared towards the easy understanding of it by LLMs, with documentation possibly including a fine-tunning dataset with which a model can be tested for proficiency in using that particular tool (like OpenAI Evals). Software will be written to be used by humans through LLMs because humans will code in natural language, and not in the language of your interface.

+1

I don't think that's so clear. When you train a deep learning model you are making it extract the gist or insight of many works and then use that pattern to produce new works. While the NN does not experience the work like a human it is definitely not memorizing.

A silly example. Making GPT write a rap battle between Keynes and Mises goes beyond a performative remix, it is transformational work, nothing is copied explicitly. If a human were to write it that would not violate copyright.

I think that to tackle this we need a new lens other than copyright in the long term.

Automate the boring stuff with Python