Healthcare is not expensive for lack of huge government-entangled corporations monopolizing and commercializing patient personal information. At least not in USA, where healthcare legislation is written by and for the industry (e.g., https://www.theguardian.com/commentisfree/2012/dec/05/obamac...), and total healthcare expenditure is somewhere around double that of most other first world countries with more universal healthcare systems, for worse outcomes in many objective measures (at least of overall social health, I realize the standard of care in America is probably second to none if you can pay for it).
I'm not sure. I've heard people handwave about this before, but if health industry corporations and their congressmen and senators are going to argue that, it should be explained and funded and voted on explicitly, rather than the alleged massive indirect subsidies via overpriced medicare and medicaid government expenditure if they're going to claim that's somehow the way health corporations fund their cutting edge R&D and high end treatments and clinics.
> There are still plenty of expensive threatments like cancer were the diagnosis is also costly just due to the tech used.
> A proper MRI is expensive, the analysis is slow and need experts etc.
I'm sure that's very true. A company that's struggles to keep a rudimentary chat app running for any length of time, spectacularly miscalculated its fiber project, to name a few obvious ones, does not instill any confidence that they would be the one to improve this. The medical industry is far more conservative and far more complicated than what Google has proven it is able to deal with IMO. You can't just move fast and break things, you can make shit up as you go along, you can't invent your own standards.
> There are plenty of things to optimize.
They've also not performed impressively on any kind of "AI" related thing, if that's what they're thinking. Their self-driving cars are still a curiosity and a long way off being revenue positive, if there is even a path to it, for example. Industry and legislative inertia and baggage aside, I don't even think a big old cumbersome dinosaur like Google has the chops to come in and make a big change. Put it this way if some revolutionary new startup company had a really great idea in healthcare, you would be disappointed if Google bought them.
> total healthcare expenditure is somewhere around double that of most other first world countries with more universal healthcare systems, for worse outcomes in many objective measures
I've seen this many times but it makes a meaningless healthcare = health correlation .
America has lots of systemic problems that affect health outcomes that are outside of healthcare. Low quality fast food, food addiction, industrialized farming, corn subsides, food deserts, deceptive food marketing to name a few.
> I've seen this many times but it makes a meaningless healthcare = health correlation .
It doesn't, it's just making an observation.
> America has lots of systemic problems that affect health outcomes that are outside of healthcare. Low quality fast food, food addiction, industrialized farming, corn subsides, food deserts, deceptive food marketing to name a few.
Sure, America has many issues and many more than you've listed. So do other countries. And they all have many differences in healthcare systems. All this makes it impossible to formulate a mathematical proof. Which fortunately I was not trying to do.
So we can't really measure the precise effectiveness of healthcare systems and health expenditure, sure. But surely the burden is on the people who would to claim that 2x healthcare expenditure for worse (or not significantly better) health than many other countries is a reasonable cost efficiency, to come up with some pretty strong evidence to support them.
> 2x healthcare expenditure for worse (or not significantly better) health than many other countries is a reasonable cost efficiency, to come up with some pretty strong evidence to support them.
yes we have healthcare system that treats disease, after that fact. Americans are a sickly group of people. Lionshare of expenditure goes to treating chronic metabolic illness. These people cannot be returned back health once they have metabolic syndromes. So matter how much money you spend on healthcare there is no real way to convert sickly ppl back into healthy people. No doctor has a way to fix diabetes or thyroid dysfunction.
Again, goes back to my original comment. "healthcare" has nothing to with "health".
Externalities caused by Airbnb opposers:
1. Block people's access to tourism
2. Worsen the economy by artificially cuts demand from supply
3. Hurt tourism workers including restaurants workers, etc
4. Create segregation which is partially what nativism is about
5. Hurt people who live there and love talking to travellers
If you don't want "other people" in your sight and zero nuisance, buy the entire street. Personal preference relying on inefficient resource allocation should be paid for.
I looked into this project (and related GPU pass through projects) a few years ago. Putting the discussion and even some issue tracking into a "support thread" in this forum makes getting information really difficult. It's the same experience when you want to install LinageOS and realize you have to go through a long thread on XDA and sometimes reddit to solve any issue you may have.
I find it unsettling that people make purchase recommendations solely based on the merit of the phones given purism contributed more patches to the Linux phone ecosystem.
I don't think that's necessarily correct. You can't even order a Librem 5 if you want it right now because of silicon shortages. You can purchase a Pinephone. Ignoring that, the price differential is insane (and I doubt people who would consider the Librem 5 USA are really saying "yes" or "no" based solely on the merit of the phone).
I think it's one thing to recognize that Purism is providing patches, but that doesn't translate to a device people can and want to purchase.
The battery life mentioned in the post is surprisingly low (see my other comment above), but you're misrepresenting it quite drastically. The post talks about the battery being at 60% after 3 hours of screen-off idling, not "3 hours of screen-off battery life".
I am not a heavy user - so not much screen time - and my phone spends 8 hours a work day in flight mode which surely helps. I charge my phone once every 6-7 days.
I know the articles phones are not mainstream. I would like a mobile OS OS :-P and as a light user I am willing to make a few concessions.
To add to that, it's mostly concerning because if Purism wouldn't be funding the development, who would? I'm hoping they'll be able to recoup and keep up their investment.
I own a pinkebook pro and it can't survive a kernel compilation (which is quite a normal thing to do for this device because it doesn't run on mainline kernel and requires patching) without dying from the battery drainage while plugged in.
I get system76's coreboot offering looks interesting because there aren't many vendors shipping that feature. What's the appeal of their other products though?