Hacker Newsnew | past | comments | ask | show | jobs | submit | dlokshin's commentslogin

"Interventions" might be experiencing the effects of Goodhart's Law. I agree that I intervene in my Tesla about once every 2 weeks, but not because it's about to get into an accident, but because it's doing something I really don't want it to do. Uber drivers do things I don't want them to do, too. If I could intervene, I would.


Sure I always see these claims that "this time FSD is good", and as a 2017-2022 Tesla owner I am continually in doubt.

Last experience of FSD was in an Uber 1 year ago where the guy proudly wanted to show it to us. Worked for about 60 seconds until the third turn in stop & go traffic where it near launched into and rear ended stopped traffic ahead. Lots of beeping and human stomping on brakes.

This was, again, during a "they finally solved it" claim cycle.

You can tell how much people believe in the product by how quickly these threads get flagged to death.


Was fun to google image search "car drives into wet concrete." Humans are worse drivers than I thought!

https://www.google.com/search?sca_esv=557782243&sxsrf=AB5stB...


I've walked over wet concrete. It wasn't really signed very clearly (there were some cones on the corners, but that doesn't automatically mean "wet concrete"), it's really not that obvious unless you're looking for it, and I was late and in a hurry.

Looking at those pictures, I think loads of people could have made that mistake for quite a few of them, as it often doesn't seem indicated very clearly, and in some situations going through the wet concrete was the only way the car could go from a parked position.


If you're in a Physics masters or Ph.D. program and are thinking about dropping out because academia is not for you, please email me: david [at] TraceUp.com. We are hiring people like you.


Interesting marketing choice that all of the images are of the actual glasses, and none of the images are of what your eyes see.


From the description it sounds incredibly difficult to photograph or represent the UI in an accurate way.


45 million people died over 4 years in Mao's Great Leap Forward from starvation and execution. 10+ million people died in Stalin's gulags alone. Are you really trying to equate this with healthcare in America?


Totally, we're past these times and living in a more stable and richer world, how is it that we're still letting people die because of competition? I think it's a fair comparison in today's standards.


This added to real GDP being more than 25% higher than it was in 2000, makes this surprising in the opposite way than the article's intent (to me, at least)


Esri is completely bootstrapped, not in a tech hub, and doing north of $1B / year in revs.

http://en.wikipedia.org/wiki/Esri


Because demand is less. The wage is still dictated by supply and demand. The supply is probably less, but the demand for that labor is far less. The demand for an engineer in the midwest is higher than the demand for an engineer in Thailand because there are less time barriers, language barriers, etc.

A house in Palo Alto, CA costs more than the exact same house in Dexter, Michigan. If they were made of the exact same materials on the exact same size of land, you wouldn't expect them to cost the same either.


I don't think the house comparison is a good example - a house in Palo Alto can only be utilised in Palo Alto. A remote worker could work from anywhere and produce the same output, regardless of location.

Pay levels based on location seem like a good way for companies to "save" money by providing a convenient excuse to pay someone less for the same work.


When you factor in the cost of living it's not an "excuse".

If, after the obligatory expenses like housing and food, the employee has more after tax dollars to spend on whatever he wants and in many ways a higher quality of living (better housing, better control over crime although that's looking to change WRT California, etc.), is he getting cheated as you imply?

I've lived in the Boston area and the D.C. area for a dozen years each, and have now moved back to my smallish home town in SW Missouri ... the differences are staggering.

Look at imroot's policies: https://news.ycombinator.com/item?id=7568217

Officially the same pay, but the people in California get an extra 1K/month for their higher living expenses. Heck, the Federal government does this for civil servant compensation.

Flip side: if the company is nasty, it'll tie up it's outside of California employees with non-competes---although that could be a bit tricky, given that it would get no support in the California state courts. How would you establish jurisdiction in the employee's home state?


That is correct. I, living in Dallas, am willing to sell my labor in Dallas for a lower price than I am willing to sell my labor in Palo Alto. Thus, a company in Palo Alto can save money by hiring me in Dallas to do the job. There is nothing inconsistent about this.


Price should be based on the value you are providing to your clients/employer. Not your expenses.


If only the price of my house was based on the value I provide my employer.


What does this have to do with anything? The amount you paid for a house is based on it's value to you. Same as your employer and it's employees. I live in the midwest - a cheap place to live. But I charge SV rates because of the value I provide my clients. I don't have to charge "less" because I live in a cheap area. It is irrelevant.


I meant it in this sort of comparison: if you uproot the house, and drop it in Palo Alto on the same size plot of land it's worth more. If the programer gets up, and moves to San Francisco, his labor is also worth more. Same person, same house, different demand (and supply).


Went to Harvard, had a bunch of friends and interacted plenty with people at MIT and would be happy to jump on the phone with you and give you a more in depth view of my experience. Email is in my HN profile.


Your email doesn't show up in your profile unless you add it to the about section. :)


:0! Thanks. Added.


The irony is that if we were to have a single payer or other form of universal healthcare like they do in France, this little girl likely would have died. In France they use "Statistical prognostic" approach which in these sorts of cases withhold NICU services for infants with statistically low outcomes.

Edit: couldn't find data for premies at 25 weeks, but at 37 weeks and below, survival rate in the US is higher than European countries. http://www.cdc.gov/nchs/data/databriefs/db23.htm


The irony is that France didn't even allow most withholding of heroic neonatal care until 2004+ (e.g, la loi Leonetti, etc), whereas the US regularly refused such care until 1984 and then enacted statutory allowances for refusing futile neonatal care in 1986 (Baby Doe revised). Also see the Texas Advance Directives Act of 1999 and withdrawal of futile neonatal care pursuant to it.

In fact, France still does not allow withholding of nutrition and fluids from an infant not on a ventilator, a practice that is permissible in the US when care is deemed futile.

If you want an example of a state that practices more aggressive withholding of care in the NICU, look at the Netherlands, which has a private insurance model more akin to the US. According to the Nuffield report on neonatal medicine, "In the Netherlands, a consensus had been reached by 2003 that ... Dutch perinatal centers should not normally resuscitate and treat extremely premature babies born before 25 weeks of gestation, because of poor outcomes."

According to the report, the difference between the US model and the Netherlands could be seen in that twice as many premature infants born in the US survived to the age of two, but in the US five times as many premature infants had disabling cerebral palsy.


Where does this myth come from that universal healthcare systems like the European ones deny coverage for black swan events? They don't. What procedures are covered has little to do with whether the system is single payer or not.


I have a relative who works in a NICU. A 1 in 3 chance of death is actually pretty good odds (relatively speaking) for a premature baby in intensive care, so no, socialized medicine would not have killed this child.


Are you sure about that? France's Infant Mortality Rate is lower than the US.

France: 3.34 US: 5.90 (Per 1000 live births)

Source: https://www.cia.gov/library/publications/the-world-factbook/...


The assumption here being that a live birth is as likely in France as it is in the US. What we should be looking for is percentage of survival for infants born sub 25ish weeks or something like that (I don't have that data).

UPDATE: And as we see from this paper from the WHO, live birth in the US is not what it means in France, for example. In the US it's any sign of life, in France and Belgium it's living for some period of time: "it has also been common practice in several countries (e.g. Belgium, France, Spain) to register as live births only those infants who survived for a specified period beyond birth"

http://www.who.int/bulletin/volumes/86/6/07-043471/en/


Those statistics aren't exactly Apples to Apples between countries. There are just too many ways countries can cheat.

For instance, some countries (like France) ignore births before x weeks whereas most countries (including the U.S.) include all births.

Beyond that, the U.S. leads the world in premature birth survival despite also leading the world in assisting people with fertility problems to become pregnant and add to the statistics that on the surface seem problematic for the U.S.

http://www.cdc.gov/nchs/data/databriefs/db23.htm


Even your own data completely undermines your point. All those other countries have very comparable preterm infant mortality rates when measured against the US. You say the girl "likely" would have died in France, but in Norway and Sweden (two of the most socialist countries in the world) the girl would have had a better chance of surviving at 25 weeks. By your own data that is supposedly proving the ineffectiveness of single payer healthcare.

Go back to Fox News where people don't actually care about facts if you want to make sensational arguments with zero evidence.


Except for the part where this is 37 weeks and not 25 weeks which is 30% of the entire duration of pregnancy. The story is from AOL is at the 5 month mark which means it could be even earlier than that.

Can't go back to watching Fox News because I don't have a TV. Nice, red herring, though.


.. and in the universal-healthcare UK, my brother survived being 12 weeks premature.

Do you have a cite for the French system, specifically that withdrawal of care in NICU? English or French will be fine.


In the UK they give services, and reasses on an ongoing basis, so different than the French. Added some info for under 37 weeks, where the US comes out on top.


So are you admitting that your original point about single payer healthcare is totally BS? Because the UK has single payer and apparently you think they do a decent job with this situation.


My first comment was directly pertinent to the French model, but same for UK--US survival rates are higher for sub 37 week old babies.


If you had government mandated "socialised" health care like the majority of the developed world, the actual cost to provide health care would have been nowhere near as much.

I cannot fathom that people accept "12 to 18 $K" as a reasonable amount of money for health care.


Got reliable sources for that claim?


> in these sorts of cases withhold NICU services for infants with statistically low outcomes

Do you have any source for these claims?


It's not super recent, but this paper discusses attitudes in the U.S.:

http://pediatrics.aappublications.org/content/120/3/519.full...

It sounds like 23 weeks and 500 grams are (were in 2007?) important benchmarks.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: