>I would argue that their sole value is to make insider knowledge accessible quicker to the general public.
Is anyone using AI to track these audacious and large bets? Seems like you could actually do this to tell which ones are insider info and which are just stupid random bets?
I would just count which side of the bet has more bets above the median bet size with a zero or near-zero bet history. Insiders are more likely to use throwaway accounts than are non-insiders. Spoofers may however deter such analysis. AI may not strictly be needed for analysis.
Just because an insider bets on something does not in any way make the bet obvious to onlookers until the event is materialized for all to see. There is absolutely no surfacing of intelligence here because big bets are on both sides, winning and losing.
Many years ago, some dial-up providers in my city offered free public logins to use their websites (for scratch card activation, account renewal, user guides, and so on). Some companies also paid ISPs to have their sites and services accessible in similar fashion for promotional reasons.
At a certain provider, all those free logins used the same firewall configuration to only allow traffic to those free services and ISP site, probably for simplicity, so all of them were accessible with any promotional login. Most of them were not useful (to me), but different agreements with ISP resulted in different call time limit until hang-up, 10-15 minutes instead of 3-5.
However, the main treasure was the addition of external page translation service as a feature on some big site. Back then, it was strictly static and server-side, URL in request gave you its HTML source with translated text strings and absolute paths to external resources, so in order for translation to work, users needed to be able to access that third party server, too. Obviously, if you gave it any other URL, the server would also grab it to translate (and choosing least similar language in parameters would leave most of the page text intact).
You can imagine that having a browser supporting tabs and switching media off was very handy for loading as many free web pages in text only form as those dial-up sessions allowed.
Obviously, WWW-to-email services for people who only paid for mail server access had existed even before that.
I don't think there are any net neutrality laws that don't exempt things like in-flight Wi-Fi, where the upstream is so heavily restricted that providing balanced services to everyone is basically impossible or leaves the entire connection useless.
With Starlink things may be looking a bit better, but I think demanding net neutrality on in-flight satellite internet and plane-to-cell-tower internet is excessive.
Thank you for sharing, still wiping the tears from my eyes. If there was ever a reminder to (try) enjoying the small, “mundane” moments of life every day it’s the part below.
“There was so much I still wanted to do. I fantasized about the smallest, most mundane things—waking up in bed, getting in my car, waiting at a red light, grabbing coffee, working. I wanted to do it all again, every day, forever and ever.”
You die from liver failure. Which is a horrible way to die. You’re bleeding and clotting at the same time. Your abdomen swells with fluid. Then your legs and your whole body. Your skin turns yellow and you’re itching constantly. You become increasingly confused and violent. Infections start brewing in that pool of fluid in your abdomen.
You should sign up for Nautilus. Monthly compact discs mailed to your door containing 650MB of assorted interactive multimedia. It was the original broadband (with very large packet size.)
It’s a bit sad that there’s nothing in this post about actually wanting to improve patient care, but instead make money for hospitals and “enhance metrics.”
It's called upcoding. It's been happening with or without AI or SV's help for well over 40 years. [0]
Interestingly, whistleblowers can recoup a big chunk of the recovered funds in a False Claims Act suit when medical providers do this to defraud Medicaid / Medicare [1]. Could be a lucrative long-term play if you get employed at a company whose sole purpose is helping hospital chains bill for care that an AI thinks may have been provided.
American healthcare in a nutshell. All of these startups in the healthcare industry focusing on "upcoding", "dealing with insurance" would go bust if private health insurance no longer existed or at least existed in a much much smaller role.
I can relate to this reaction for a variety of startups. I still am interested in working for these light mission or neutral mission startups, rather than crypto or something empty, but I wish we had a lot of deep mission startups to choose from. I blame the diluted mission problem on VC + hypergrowth. If more startups could focus on less aggressive capital and growth goals and let the founders dream more, I think we would all be more excited.
Our models don't only suggest up-coding possibilities, but also chart quality improvements! As a result, patient's charts become more accurate, allowing for potentially better patient care in subsequent visits.
Also, the hospitals being able to bill more accurately allows them to have more resources to invest in patient care, and when you consider that a lot of hospitals operate on 1-3% profit margins while health insurance companies operate on a sometimes exponentially higher %, it's actually one of my favorite aspects of what we're building here.
Mind if I ask a very non-defensive "how come"? The logic in the comment you're responding to is legitimately the way I think about one of the impacts our products have re: patients.
