1,991 karma · joined October 24, 2013
Biohacker, developer, nerd.
[ my public key: https://keybase.io/jackkingsman; my proof: https://keybase.io/jackkingsman/sigs/KkmLhjehzVC4pTPirRSxU30diNt68EfKvEkQOV9UOvA ]
Crazily, I only stumbled upon this because I ordered some discount blood labs and the requisition had Health Gorilla on the letterhead, which I found an absurd company name, so I googled them, and found the lawsuit which was filed the day prior. Absolute chance.
Essentially, Epic, a massive healthcare company running the majority of electronic healthcare/medical record systems for hospitals/etc. makes data available to various data brokers, who then subcontract to other healthcare providers. The goal of this subcontracting is that if you e.g. come into an emergency department unconscious, but with identification, doctors can pull data from the broker, solemnly swear that they're treating you, and gain access to your whole medical record. Generally, good actors in this space will seek signed consent paperwork, or have policies in place with narrow carve-outs for emergency access, but there is (to my understanding) not a centralized, standardized system of access request, patient approval, and auditing.
There have been many issues in the past with shady providers who are, indeed, treating the patients, also turning around to sell the data they have to legal firms looking for plantiffs for lawsuits under the guise of "we're helping the patient by potentially giving them access to lawsuits that will advocate for them."
This current lawsuit alleges that the data brokers this time were simply turning a blind eye to completely fraudulent actors who never had the patient under their care, and that their access was knowingly used to bulk-mine patient data for lawsuit opportunities.
I love making friends with strangers, but usually rely on the "handshake protocol" of a casual observation or small talk that is then accepted (with a similar slight-deepening or extension of the thought) or rejected (casual assent or no response at all), until the bandwidth opens and I can foster a more meaningful moment of connection with a pivot like "Oh awesome that you do $THING for work. Do you enjoy what you do?" or "Oh I don't know much about $LOCATION_YOURE_FROM. Good spot for a vacation, or good spot to drive straight through?"
As somewhere between "thinks like an engineer" and "on the spectrum," I really enjoy hearing others' strategies or optimizations (optimizing for quality, connection, warmth) for social situations.
It used to be a general rule of thumb that you could build a computer of roughly equivalent power for the cost of a game console, or a little more — now the memory costs more than the whole console.
HIPAA's scope is actually basically nonexistent once you get away from healthcare providers, insurance companies, and the people that handle their data/they do business with. Talking with someone (even a company) about health conditions, mental health, etc. does not make them a medical provider.
Both my partner and I used tablets for notetaking through college and found it at least analogous if not superior to handwritten notes, since it became easier to link topics that you might otherwise need to back-reference on paper that could instead just be a big arrow. Lots more freedom to use arrows, visual linkages, and asides when you weren't constrained by 8.5x11 paper (which maybe allows a bit of that but otherwise forces linearity, more or less).
One of my friends sent me a delightful bastardization of the famous IBM quote:
A COMPUTER CAN NEVER FEEL SPITEFUL OR [PASSIONATE†]. THEREFORE A COMPUTER MUST NEVER CREATE ART.
Hate is an emotional word, and I suspect many people (myself included) may leap to take logical issue with an emotional position. But emotions are real, and human, and people absolutely have them about AI, and I think that's important to talk about and respect that fact.
† replaced with a slightly less salacious word than the original in consideration for politeness.
Where can I learn more about that? My google fu is failing me.
I wouldn't say /never/ attribute to malice anywhere we're in the vicinity of an an enormous data actor with a not-great track record, but probably at least /even/ are the number of cases of privacy violation attributable to maliciousness vs. terrible design that is either excessively encumbered or insufficiently granular.
I'm perfectly happy for Acme Random App to scan for `pps-setup-wifi-**` at one time, but not all wifi networks forever.
If there was a paid-only search engine with dubious ethics practices that was overwhelming my site with traffic in order resell search trained off of (among other things) my personally generated content, I would absolute block it.
LLMs are not search engines, and I'm not gaining any followers or customers in any meaningful way because an LLM indexes my site.
> it also cuts you off from the fastest-growing distribution channel on the web.
I haven't seen the needle tip at all in my acquisition channels from LLMs. Unless you're a household name or very large, LLMs aren't going to shill for your business.
> most LLMs have an agentic web-search component that will actively generate links
Totally. Which is why I don't care if the LLMs index it. Let web content search be good, and lead LLMs to good content; product placement in LLM weights ain't what I'm gonna optimize for, or even permit, if it comes at a cost to me and my infra.
So, I'd agree -- safety info from an LLM is bad. But generally, the /flavor/ (heh) of information that such data comprises is REALLY good to get from LLMs (as opposed to nuanced opinions or subjective feedback).
And uv is fast — I mean REALLY fast. Fast to the point of suspecting something went wrong and silently errored, when it fact it did just what I wanted but 10x faster than pip.
It (and especially its docs) are a little rough around the edges, but it's bold enough and good enough I'm willing to use it nonetheless.