755 karma · joined October 9, 2014
These very practical daily occurrences can make devices incredibly annoying and frustrating for typical consumers who want it to just work.
Second, there is a standardization of codes for the most common procedures one would shop for, which is CPT. There are HCPS, DRG, and other codes as well - but they’re much less common among the types of care one would reasonably shop for in advance. A crosswalk between code sets is also possible for many procedures.
There is certainly not a standardization of the use of codes in all cases, which I alluded to in my message. It’s extremely challenging to know what codes you will receive. And placing provider malice aside, often times providers themselves do not even know what it will be before going in.
To this end, bundled episodes have become increasingly common for certain types of procedures. But it’s far from exhaustive and there’s much room to go.
In any case, the goal was never for this data to be directly used by patients. It was to expose the data, and make it available for third parties. Most of those efforts have fallen flat not for lack of data, but lack of a patients ability to reasonably engage with it for the reasons mentioned above.
You can see more about the act here: https://www.cms.gov/healthplan-price-transparency/plans-and-...
The reality is that price transparency already is federally mandated and you can get a list of every contracted rate for every procedure any doctor that is credentialed with your health plan will perform. They’re all available on every national payer’s directory in a machine readable format and updated every month. Startups have formed around making this data available to consumers. This has been the case for a while (well, the number of procedures was limited before, but the concept still existed). It made headlines years ago.
The problem is three-fold (there are multiple but these tend to be the most problematic):
1) It’s almost impossible to know, for any non-trivial visit, the actual procedures that will be performed. This makes price shopping almost impossible to begin with. Not only is it difficult to understand up front, it can (and does, with regularity) change halfway through the procedure - sometimes even after the fact if the procedure was improperly coded. To some extent this is just a result of an overly bloated and complicated billing machine, but it’s a real issue today.
2) Healthcare consumers at their most vulnerable (and, typically, most expensive) are not in a mental space to be price shopping for care. Savvy individuals might comparison shop before an orthopedic surgery, but nobody who just suffered cardiac arrest or was diagnosed with cancer will spend weeks finding the affordable doctor.
3) Healthcare is an intensely personal and emotional interaction. People don’t want to price shop for care even if they can and are in the headspace for it. You don’t want the budget doctor performing the bypass surgery, you want the good one if that’s an option for you.
This can and likely will change in the future. But today the system is broken much beyond price transparency.
More specifically, new entrants (and especially so in new categories) regularly get the luxury of extended time to iron out issues, create a better UX, find all the edge cases, invest in internationalization, distribute to multiple platforms, integrate with other systems, address security issues, and more. These tend to be exactly the kind of things that make recreating the incumbent products hard.
If OpenAI continues on its current trajectory, it can easily end up with a team equal to or larger in size than all of Google working on search.
Again, this isn't intended to be discouraging at all. Just don't operate on the assumption that eligibility APIs showing out-of-date deductibles is normal and not an edge case!
My point in my original comment, which I believe still holds, is that there is no objective truth to this number. Multiple buyers can arrive at different definitions of worth depending on how they used the asset and when they ultimately sold it. And even if they both arrived at the same value, it can only be known in hindsight. For example, say I purchase a stock solely as a monetary investment for $10 today, in 1 year the price drops to $2, and in two years the price jumps to $50. The answer to whether the stock was "worth" what I paid depends on when I decided to sell it. If I had to sell that stock in a year to buy a house, it certainly was not. If I could hold the asset for 2 years, it was certainly worth it (and then some). Or, going back to your example before, if the cash flow was less than I expected - that's not great. But next year it could be more than I expected. The time horizon is a critical component here, as are the factors unique to the individual buyer.
Your statement that the property was not worth that much can only be said with the benefit of knowing what the market is today. In 6 months, the asset may be worth 25% more than was paid for it. It's likely that's not the case, but clearly many people thought it was likely the asset would generate more money than it did when it was bought and were wrong. We may also be wrong today.
So, when we're talking about the "worth" at the market level we almost have to be talking about willingness to pay. We assume that will price in estimated future value. There's not much else we can usefully measure here.
But that's not a particularly useful definition of worth. Worth measures a moment in time. We can only price something based on what we know about it today and what we predict about it in the future, and we're not perfect predictors of the future.
I think about "worth" using the only definition of worth that is particularly helpful (to me): the amount you could get for the asset if you were to sell it now. Part of that decision almost certainly a prediction of what you believe an asset will be worth in whatever time horizon you plan to sell it. You may be right or wrong about that prediction. But it doesn't change that in the moment at which your purchased the asset, it was "worth" that to you/the market. The reality is that many, many people were under the perception that the value of these assets would continue to appreciate as they were all collectively willing to pay what we now believe to be high prices for them.
In hindsight, we may all have been wrong. But in that moment, I could absolutely have sold the asset for the price I paid to purchase it. So it certainly was "worth" that when the transaction happened.
But regardless of that, whether the values were inflated because of a market flush with cash or not, it doesn’t change the fact that crashes can and do have ripple effects.
I believe you also understood precisely what I meant, and are trying to form a gotcha.
In any case, I’ve lost the plot on your point here. That interest rates impact spending and corporate decisions is verifiable fact, using the ever-elusive scientific method which you’ve so helpfully educated me about. If you truly believe in this method, I’d suggest you exercise it and watch how interest rate changes shift economies and markets every time they happen in history. It just so happens to be precisely the exact reason why the Fed moves them in the first place. It was widely discussed when rates began rising that the impact would likely be an increase in unemployment. The Fed specifically addressed this point, and stated that while it was not an explicit goal to increase it in this case, it may happen.
Public markets operate on opinions and feelings more than science. That sentiment of the markets is not high right now is clear. Look at every publicly traded tech company’s performance in the last 12 months.
If you do understand that higher interest rates suppress higher risk investments, then there's a direct line between that and slowing VC funding and depressed investor sentiment, which is ultimately what's feeding these layoffs.
The indirect link is not always true. Interest rates can be risen in a bull market. When done slowly, this effect is less pronounced. In this case, we massively shifted rates in a short amount of time during a period of high inflation.
When interest rates drop, safe investments (such as bonds), become less attractive. The yield on those investments is not high. This means that excess capital gets reallocated towards riskier bets (such as stocks, or venture investing) to try to yield a return that way. More speculative bets happen as a result of free flowing cash to high growth organizations. Companies are willing to lose money in exchange for market share because investors are willing to bet on companies that are losing money on the off-chance that one of them yields a 100x return. This game becomes more attractive when other means of growing investments are harder to come by.
Now, interest rates are higher. I can throw money into a savings account at 4.5% APY. A potential 7% return by throwing it all into high growth stocks (which also carry significant downside risk) is comparatively less attractive. I'm now much more incentivized to invest in companies that are stable and carry less downside risk. Those companies are the ones that are spending more responsibly and generating positive cash flow.
Every single time a layoff is done, there’s a conversation about what initiatives are on the chopping block as a result. Investments that are now lower priority are cut.
It’s really quite disappointing that the prevailing sentiment on HN seems to be that all corporate leadership are a bunch of bumbling fools. Just like technology is more complex than it looks on the surface, so too are large enterprises.
For most technology companies, salaries and compensation are one of if not the top line item in the budget. And during a time of cheap money (which is objectively true - this is not someone regurgitating other comments), companies invest in new initiatives.
So when headcount is the top line item, and new initiatives are less desirable as capital becomes more expensive…it’s pretty clear that there’s a small set of viable options to make the math work again.
It’s less that I now have a specific endpoint, but rather I have a target (or multiple targets) at which I can decide if we’re on track and, if not, how I may want to change previously made prioritization choices.
If you tell me a day, then 5 days later if it’s still not done I may not rip you for being wrong but I may decide it’s time to pull the plug on that initiative since it no longer meets the criteria under which we agreed to do it.
I’m all ears for another proposal, but “just keep building stuff with no regard for how long any of it takes” is not much of a strategy and doesn’t reflect the way that decisions have to get made in the real world with limited resources.
Asking the builders to estimate the time they need is a good sign of high trust in that team. It’s healthy. We should want more of it. The business’ side of the commitment is to understand that it is an estimate and should be treated accordingly.
My point is that it may well not matter whether a thing is sentient or not if a well-trained algorithm can achieve the same or better results as something that we believe is sentient.
We seem to operate on the assumption that sentience is "better," but I'm not sure that's something we can demonstrate anyway.
At some point, given sufficient training data, it's entirely possible that a model which "doesn't know what it's saying" and is "stringing words together using an expansive statistical model" will outperform a human at the vast, vast majority of tasks we need. AI that is better at 95% of the work done today, but struggles at the 5% that perhaps does truly require "sentience" is still a terrifying new reality.
In fact, it's approximately how humans use animals today. We're really great at a lot of things, but dogs can certainly smell better than we can. Turns out, we don't need to have the best nose on the planet to be the dominant species here.
The point is that, however, you can pick almost any objective health measure and see a similar result. So either all of the measurements are wrong in favor of other nations, or there's a clear problem here.
I think the bigger challenge here is that health in a nation is about much more than just care delivery services. The nations that do better than the US also tend to have significantly greater, and more accessible, social services to the population. Lower income households will have access to higher quality foods, for example. Healthcare services can be rendered even to those who are not actively employed. etc.
Effectively, the US will need to decide if it cares about the health and well being of its population. If it does, we have just about one of the worst ways in modern society to accomplish it. Our method is more expensive, less enjoyable, and has worse outcomes.
In fact, most developed nations do. But part of what makes it so difficult to have a productive conversation about healthcare is that everyone has experience with the system in every single country, and bad experiences get passed around. So we say X country is doing better, and someone chimes in explaining how that country has its flaws.
They all have their flaws. But the US health system is measurably worse than most other developed nations across most axes that we care about. Though there are interesting questions about how much worse per-capita healthcare spend actually is given the comparatively higher R&D investment in the US.
Quite simply, there absolutely is neither enough time nor mental space to deeply evaluate and understand these types of risks for all of your vendors. Further, this is one of those risks that is abundantly clear in hindsight but a person without deep financial experience or one who hasn't gone through this exact problem would really have no reason to understand or ask about it.
In fact, a person without intimate knowledge in payroll software might have no reason to even understand that there is an intermediate bank account in which funds are parked en-route to employees. It's just extremely unrealistic to expect someone to be thinking about this when picking a payroll vendor. Just like it would be unrealistic for me to expect an employer to perform a deep analysis of an insurance company's historical MLR and books to ensure they're sufficiently liquid and solvent to fund an employee's medical emergency.
Like it or not, social proof is a fairly critical construct in the world, and without it we'd be stuck spending more time analyzing our decisions than we do living with them.
The fact that a bank failed in 24 hours and Rippling made changes to the system to route around that bank and raised capital to front customer payrolls on behalf of their employees is incredible. The default action would be to wait it out.
You really should be able to show appreciation in a situation like this. This is not “normal course of business” stuff.
This is not a statement about those individual people’s skill levels, but simply a reality of the constraints applied to them by their organization (largely following prewritten scripts)
It is, technically, true. But it’s not very helpful in understanding why these things happen and which are good vs bad.
It indicates that there’s still something a human does that the machine doesn’t, even if we’re not able to place what it is. This is neither an argument for nor against progress towards AGI, just an observation. It’s interesting regardless (to me).
There are open questions about whether or not it really needs to reason given sufficient training, but that seems to be the gap here between the human and the machine.
In any case, it's an interesting exercise regardless of your opinion/stance on the matter!