Yeah I don't buy that part either. The problem is inherently hard and pushing the responsibility for policing content to a corporation is not okay. As a specific example, I was listening to a podcast I was listening to earlier that was dealing with how Facebook filters content (I think radiolab). I'm para-phrasing heavily since it's been a while since I heard the show, but I think I got the gist of it right.
- There was a violet attack in Mexico against a journalist by some drug cartel and a random body part (leg or an arm or a head or whatever) was chopped off and posted on social media by the cartel. The people protesting cartel violence picked this image up and were using this image as a part of their protest. Facebook's censors allowed it until the image popped up on some school kid's feed in America / England and all sorts of outcry later, it was removed from the site as a violet image.
- The boston marathon bombing happened and gruesome images of people lying on the floor with limbs strewn about started getting shared on the site. Facebook's policy at this point specifically said no violent images on the site so they got blocked by the censors. Media picked this up and raised an outcry on how facebook was censoring these images and basically someone high up said 'leave it up or else' and the images were allowed on the site.
This is clearly hipocrisy and there's no right answer here. Traumatizing school children with violent imagery they didn't even want to see is not great while the Boston marathon bombing was a significant political event in America and those images didn't deserve to get silenced.
This ends up being a question of free-speech and what sorts of content Americans were okay with. There are no right answers and I believe the govt. definitely should step in and provide guidelines here.
The disingenuous thing is using this as an example of how Facebook is okay with govt regulation while resisting any regulation around things that can hurt them like any number of their privacy mishaps, shady 3p data markets and ad tech in general.
>This is clearly hipocrisy and there's no right answer here.
There is a lot of hypocrisy going on, but this doesn't mean there is no right answer or that the right answer is to invite government fucking censors to control everything.
There are pretty obvious things platforms can do:
1. Instead of opaque algorithms, give people control over what they see.
2. Either do bare legal minimum of moderation or create clear, exhaustive, stable and unambiguous rules for which content is and is not allowed on the platform. Once the rules are set, take a stand and abide by them in all cases.
> Beijing has been developing a nationwide facial recognition system using surveillance cameras capable of identifying any person, anywhere, around the clock within seconds.
So before all the privacy activists are up in arms, this is pretty incredible and it looks like they're getting pretty close to eliminating all violent crime. I think that's an incredible achievement if they can pull it off. In an even broader historical context, individualism and capitalism have had their run for 100+ years, maybe this is the rise of a new ideological movement.
There are many theoretical ways to eliminate violent crime that would be to most people's eyes vastly worse than letting crime continue, so it's dangerous to elevate one metric above all others.
Here are some theoretical ways to eliminate all crime which would likely be worse for everyone:
- Eliminate all people
- Keep all people physically separated from each other.
- Remove all freedom of expression or free will from people.
Those are, obviously, some of the most extreme possible ways to accomplish that goal, but it does illustrate that there's an obvious trade off being made, and that even in cases where it might not be as obvious as these, we should identify and think about the consequences of that trade.
> So before all the privacy activists are up in arms, this is pretty incredible and it looks like they're getting pretty close to eliminating all violent crime.
To be fair, Chinese people anywhere in the world have exceptionally low rates of committing violent crime, and this has been true since long before anyone started a facial recognition program.
Since nothing a government does is a crime, by definition, then by consolidating all crimes, violent and otherwise, to the government then we have eliminated all evil! Brilliant!
> So before all the privacy activists are up in arms, this is pretty incredible and it looks like they're getting pretty close to eliminating all violent crime.
Even if it's successful, it will only be used to eliminate the crimes committed by people without connections. Powerful and connected people will be allowed to get away with crimes.
Violent crime rate is very low because of another reason: Security cameras everywhere. Criminals are either already in jail because of more efficient policing, or stop doing stupid things after watching the real stories on TV about how criminals were caught.
You mean, if it’s real and even if it does work that well, besides the state sponsored kind?
China has millions of people in internment and retraining camps...
Regarding your “new ideological movement” comment. It’s not really... we saw the same ideology (to a less effective level) in all dictatorships and communist countries. Secret police, thought police, etc.
We have millions of black people incarcerated, the result of systematic discrimination over generations.
I'd also like to point out your statement doesn't really add to the conversation as any people in prison, or considered an enemy of the state can be spun politically.
Give how important this subject is, it's important to not invent facts.
Your core statement is false and off by at least four fold. There are roughly half a million black people in prison at all levels, not millions. That destroys your setup by debunking the extreme exaggeration. However given it is an important subject, we shouldn't stop there, let's examine the situation.
There are around 1.2m-1.3m people in state prisons, 500k-600k in local jails and 200k in federal prisons. Roughly 1.9-2.1m total across all jails and prisons at all levels. Several hundred thousand of those people are processing through the system, awaiting trial, etc. at any given time.
Roughly one million of those people are there due to rape / sexual assault, murder, assault or robbery.
As of 2016, 339,000 hispanics were sentenced to either federal or state prison; 439,000 whites; 486,000 blacks. That's from the Bureau of Justice Statistics (no figures for local jails). Over the last decade black people saw a decline of roughly 120k people in prison at those levels, white people saw a decline of about 60-70k, hispanic figures were steady.
Can you explain what part of society forces people to rape and murder - whether hispanic, white, or black? I grew up poor, most of the people I knew growing up were relatively poor or very poor, many with broken homes and absent parents, I'd like an education on this. I never saw anything related to social oppression forcing people to rape and murder.
There are a lot of societies with enormous oppression, corruption and poverty, which lack high murder rates and high violent crime rates. China is one example of that and there are several others in Asia. Eastern Europe has also had many examples of that over the last several decades.
Russia has a murder rate about ~8-12x worse than China along with far higher violent crime rates, despite the two countries being at similar levels economically per capita and both having oppressive political systems. What's the difference? Russia has a much worse culture of tolerating and encouraging violence and murder.
I've known a lot of poor people, the only difference I've ever seen between poor people that were violent and poor people that were not, is culture and it was always a choice.
>China is one example of that and there are several others in Asia.
Keep in mind that China has a heavy incentive to lie about their prison population and/or play games with the definition of "prison population"; after all, it's important that a harmonious society has few people who break laws (implying a low prison population). Eastern Europe under Communism faced a similar cost/benefit structure.
Therefore, as they have an incentive to cover that up, it's safe to make the assumption that the real numbers are likely significantly higher.
If you grow up in the ghetto you don't have the time to think about college prep, violin lessons, or things like that traditionally needed to get to college.
Maybe your parents were incarcerated, and maybe your parents were incarcerated because of a lack of opportunity. These things stack up over generations until the point where it becomes normal for that part of the population.
Yes, many may be incarcerated for good reason, but you need to understand why. For this you'll either believe that black people are simply more likely to be murderers and rapists or there's another reason why.
I’d assume it’s because he’s also replying to an exaggeration. The GP claimed “millions of people,” but the US State department says it’s around 800,000 that are in these de-radicalization centers.
My point is criminal behavior is defined by the ruling class. So this pervasive 24/7 monitoring doesn’t necessarily reduce the overall harm to a society - in fact it may make it worse.
Also:
> We have millions of black people incarcerated, the result of systematic discrimination over generations.
Not even close to the extent China is persecuting - today. I won’t speak for generations past, clearly not ideal. But I think that’s kind of the point, we agree it’s not ideal.
>Not even Close to the extent China is persecuting today
You might want to look that up. America has the highest incarceration rate in the world, even after you factor in the worst estimates for the Uyghur reeducation camps.
Incarcerations are not persecutions. Arguably a negligible number of percecutions occur in the U.S., as most of the incarcerations are due to violent crimes. Most of the Uyghur re-education camps (which is just one of many examples known and unknown) are not due to violent crimes, but ethnic or political backgrounds (I.e. persecutions)
There is a strong legacy component to African Americans and violent crime. You either believe that they are simply more inclined to violent crimes, or there's societal factors out of their control that push them towards that, such as denial of opportunities or institutionalized discrimination.
These types of things have long lasting effects to this day, that affect the African American population. Saying that they are now committing violent crimes so that it's justified is a copout on the factors that lead to it.
For example prestigious universities have strong bias towards legacy applicants. If your forefathers were incarcerated by discriminatory laws in the past, you're unlikely to be a legacy admission, or benefit from any legacy policies in society.
In America activists can potentially sue/appeal for to free those wrongly imprisoned, and the incarceration rate is about 0.75%. The Uyghurs, by contrast, are currently imprisoned at a rate of minimum 6% (calculated via lower estimate of concentration camp size and upper estimate of Ughur population in China), another 6% of their population is matched by live-in Han Chinese family monitors, and no one is even pretending that they broke any laws.
Yes but African Americans make up 34% of the 6.8 million that are incarcerated. Maybe the rules causing this disparity are not written down anywhere like in China, but it's certainly ingrained into American society in practice.
I mean that's the thing, China is going to implement it, we get to see what happens, and I don't think there's much we can do to stop it. I'm interested in seeing how historians look at this period of human history since it will theoretically be so objectively (? maybe? At least plentifully.) documented.
Either it's documented and the opposition can make educated criticisms, or it's not documented and the oppositions is politically motivated. You can't have both.
Is there something that tracks the number of minutes a daily active user spends on the site? I still login to facebook but I barely post on it anymore and spend way less time than I used to a couple of years ago.
Needs to be adjusted for cost of living, salary changes over time by profession (Tech salaries are in a bubble rn. which makes 150k reasonable by bay area standards, but they may go significantly up / down in the future) which is complicated enough but there's edge cases like Universities / Non profits etc. that can't pay salaries to compete with the private sector but people work there anyway because of passion.
I disagree, if you can find cheap programmers in Des Moines, Iowa, then by all means expand there, spread the love within the country before going for import labour. As for bubbles, that's the point, there shouldn't be any if labour is available and willing to come in.
Most IT / h1 roles are not based out of the Bay Area [1]. Significant amounts of H1b population goes to doing cookie cutter consulting projects spread across the country and setting a 150k min on the wage means that you won't be able to hire H1s at all in other parts of the country. Obviously this creates a lot of demand for American programmers and personal experience suggests that this will cause a labor shortage.
That doesn’t seem too hard to solve. You just say that an h1-b job needs to be at the 85th percentile, salary-wise, in a basket of similar jobs by location/position.
That's what the regulations say right now - for each title, there's a prevailing wage level that DOL publishes and a H1 has to meet that wage level. It is however a single level across the country so it tends to be a bit on the lower side.
Actually it isn’t. “The prevailing wage rate is defined as the average wage paid to similarly employed workers in a specific occupation in the area of intended employment.“
https://www.foreignlaborcert.doleta.gov/pwscreens.cfm
I don't think that's necessarily the case. A lot of people that work at a university do so because they're an international grad student who has no other way to stay in the country. They're barely scraping by, hoping they can finish and that an advanced degree will land them a job.
Oh man ignoring the privacy implications of this, all the "product" people at Facebook are going to destroy WhatsApp as we know and love. It is going to become a giant monstrosity with a 500MB binary size, lag, whole bunch of tracking code and super slow servers. It has begun to a certain extent already and it's only going get worse.
I assume they think that the network effect is going to lock users into WhatsApp but the moment it becomes too painful to run on a 100$ Android phone with 1GB of RAM, it will inevitably die. Sure it's not going to be instantaneous but I'm a 100% sure that all the PMs that run Facebook Messenger are itching to get their hands on WhatsApp.
I understand these changes are only on the server side, but I imagine the client side is not too far away. Some client changes are inevitable because I'm pretty sure they'll build a "unified" API for all these apps and it's is going to contain a whole bunch of messenger service code (because look at all those messenger features that noone cares about, surely we can't just drop what a whole org has been working on for two years)
> all the "product" people at Facebook are going to destroy WhatsApp as we know and love.
Well, let it just die quickly! There's not much to love in products from a company that's repeatedly so hostile to its users. I'm all for anything from Facebook dying sooner, and will be cheering when that happens.
You don't trust Zuckerberg to keep your private messages private? Even if that somehow happens, he'll be selling your
messaging network info to everyone with two nickels to rub together.
No thanks. Delete Messenger/WhatsApp and move everything onto other services.
This is precisely while I'm on instagram: There's no pretense of anything being private. It's like the best parts of Facebook (cool pictures/info from people who interest me) without the bullshit (TDS-fueled ramblings).
I'll be out of there as soon as a federated alternative pops up and gets the least bit of traction.
Seconding this; The article's talking about a consolidation of the messaging infrastructure, not the product experience. In theory, this could be done with super minor changes to the product code, and in practice, there's nothing stopping Facebook from making arbitrary product changes to WhatsApp anyways.
There is not only that social form of "network effect" keeping in, but also the "free basics" network effect. I noticed that while travelling Asia: In many areas you got Facebook's limited form of free internet, where Facebook's services can be used even without money in the prepaid card. Thus WhatsApp is the central communication channel. Many shops, taxi/tuktuk drivers, guesthouse owners etc. use it as primary communication channel to make reservations etc. That lock in is strong. (One has to see how Facebook will monetize this, but we'll internet==free basics==Facebook+WhatsApp gives them a strong advantage.
I note this story is the top read item on the BBC news right now, clearly people do care quite a bit.
The web has evolved with messaging APIs and server side push notification things, how easy is it to put together a group chat for just your family these days? The wider friend-verse can stay on these common platforms but the inner circle can be kept off servers and so long as someone in the group has the messages then a complete transcript can be had?
I don't believe it can be that hard. Facebook has become a behemoth and you don't need all that to just share messages within a small family sized group. Whilst Facebook merge their triplet of behemoth apps I am sure it can be possible to put something together that just works for the inner friend/family group everyone has, a little advert free zone that just does text and phone recorded media. Can it be done utterly serverless or is this something the blockchain crowd have solved already for me to have dismissed as snakeoil?
> I am sure it can be possible to put something together that just works for the inner friend/family group everyone has, a little advert free zone that just does text and phone recorded media.
I've been researching this lately and I'm about to test out private (but possibly federated) instances of hubzilla and nextcloud as soon as I can get some time.
Both looks extremely promising although I know there are issues.
> The driver might almost forget the help he’s getting, and attribute the success to his own powers
I really like this idea but I feel this is the key in getting it right. Several video games achieve this feeling of seamlessness when translating raw input from a controller to in-game actions. However this does occasionally go wrong and can be a frustrating experience in video games (especially if you're playing online competitively) but in the real world it can have much more drastic consequences.
Also, given the complexity of the real world, it's somewhat scary to think of a scenario where you're actually trying to take a drastic action on the road but are prevented by the system from doing-so because it feels that the action is unsafe.
The lane assist in my VW can get a little troublesome in areas of road construction where lane markers have been painted over and such. At times the car picks up the wrong lane area and encourages me to stay there instead of the real lane. It’s more of an annoyance than anything but it’s still a distraction when at highway speeds
I once had lane assist on a rental car (so one unknown to me) hide that one of the front tires had a flat until things got too bad to hide and then suddenly give up and dump the problem on me. Luckily no accident, but I'm certain that on a car without electronic support I would have felt earlier that something is odd.
Next version should detect whether a site is under-construction (# of cones per 100m, GPS-tagged construction zone from publicly available DB, sudden speed change, unexpected traffic etc.) and turn on the soft-mode for lane-assist where it says 'You might be driving through a construction zone, so lane assist will not fight'.
I feel that until a car can ascertain that it’s in a construction zone using only vision (like you and I do) we’ll never achieve what we need. Any autonomous system that relies on (or even incorporates) stored databases of any kind is implicitly worse than a teenage first time driver and should therefore not be considered acceptable.
Indeed, we don't get basic things like GPS signal 100% right 86400 seconds per day. Relying on some remote DB which can crash, be hacked, cut off from internet, down for maintenance etc will be a problem. Or car just losing data connection (like it happens with phones).
Its nice to dream about ideal future, but if lives rely on reliability, it either works 100%, or shouldn't.
100% is stronger than we need - humans are not 100%. If we were there would be no crashes. Even sudden mechanical failure could just affect the one car and not cause others if humans were perfect. We just need to be better than humans.
I often see the argument that you might need to floor the throttle and swerve pulled out in these threads but I'm dubious that this is a real problem. My gut feel (and no, I don't have data to back this) is that far more accidents are made worse by people deciding to use that response than would be made better by the car overriding that decision and putting the brakes on to come to a somewhat controlled stop.
I'm sure there are certain types of crashes that can be avoided in that way, but I trust a collision avoidance system to make that judgement far more than I trust a human driver who was only half paying attention to the road five seconds ago.
Speaking of video games. They need to get this system setup for race tracks like Nurburgring where the car doesn't let the driver go to hot into the corner by controlling the throttle response and braking. Also, control steering to prevent spinouts including auto counter steer.
What is the point of a race track if you even further abstract out the driver skill than we already have? (asks the person that races a fairly analog car)
You're never going to get skilled if you don't learn how to actually drive the car in the first place. A car that 100% sticks to the road with no slip is not as quick as a car that is being pushed to the point of carrying a slight slip angle through corners, and you're not going to find out how that works with constant intervention. It is about carrying speed through corners, not negotiating them with the help of some electronic aids and pinning the gas down the straights.
If I could show up at the Nürburgring and hot lap a 911 on nanny driver mode I'd do it in a heart beat. Until then I'll just stick with not getting into Berghain.
Speaking of DLCs, didn't Porsche sort of offer an hourly "Performance Package" for their cars that could be activated OTA. Or I guess the interviewed guy was just spit-balling ideas...
It's like go-carting. People like to race cars without having to be skilled or deal with the risk of being in a car. Most cars already have a bunch of electronic nannies(eg: stability and traction control, abs brakes) to prevent people from spinning their car out into a tree. Also, nannies to prevent them from ruining their car like rev limiters.
Developing skill is hard, buying a product is easy. What you're saying makes total sense if you want to become an excellent driver but most people just want to buy some related equipment and watch events from their couch. Nothing wrong with that if it makes them happy, though I think that advertising pushes people towards making less fullfilling choices.
People enjoy roller coasters. It's not a big stretch to imagine it being fun for plenty of folks to whiz around a race track with some of the work being handled by a computer.
> However this does occasionally go wrong and can be a frustrating experience in video games (especially if you're playing online competitively) but in the real world it can have much more drastic consequences.
Even a manual override would be insufficient because by the time you realize it won't behave correctly, it's too late to apply the override and then take corrective action. I would probably not trust this system, in my hubris.
Presumably that means less jobs in the UK and more jobs in the EU to manage that wealth for one. I'm not sure what that means for future growth of the UK financial services industry.
I have to say, every time I come across this book, it seems like bullshit not backed by any science at all and yet it gets rave reviews on HN. As a one-time thing, I can put it off towards some propaganda campaign ala Tim Ferris / The four hour body but this is a pattern. Don't know what to make of it really.
- It sets off every rational person's bullshit detector.
- It works anyway.
- The effort required is minimal (just reading the book is enough for many people).
- The payoff is life-changing.
When you experience all this first-hand (as I did), it's hard to not become an evangelist. It has no business working, and it yet it does. In this community, I think we have a strong desire to share bewildering things like that with each other.
- There was a violet attack in Mexico against a journalist by some drug cartel and a random body part (leg or an arm or a head or whatever) was chopped off and posted on social media by the cartel. The people protesting cartel violence picked this image up and were using this image as a part of their protest. Facebook's censors allowed it until the image popped up on some school kid's feed in America / England and all sorts of outcry later, it was removed from the site as a violet image.
- The boston marathon bombing happened and gruesome images of people lying on the floor with limbs strewn about started getting shared on the site. Facebook's policy at this point specifically said no violent images on the site so they got blocked by the censors. Media picked this up and raised an outcry on how facebook was censoring these images and basically someone high up said 'leave it up or else' and the images were allowed on the site.
This is clearly hipocrisy and there's no right answer here. Traumatizing school children with violent imagery they didn't even want to see is not great while the Boston marathon bombing was a significant political event in America and those images didn't deserve to get silenced.
This ends up being a question of free-speech and what sorts of content Americans were okay with. There are no right answers and I believe the govt. definitely should step in and provide guidelines here.
The disingenuous thing is using this as an example of how Facebook is okay with govt regulation while resisting any regulation around things that can hurt them like any number of their privacy mishaps, shady 3p data markets and ad tech in general.