The 9th Circus was surprisingly and astonishingly correct here (mark it on the calendar!).
I've not read the ruling, but from the summary it sounds like their hands were tied by a recent Supreme Court ruling. Give them a bit of time, they will figure out a way around that ruling and eventually decide such contracts are not lawful.
However in this case they knew that exactly 293 cars were affected and which ones they were. This made it easy to track down and remove the unsafe cars from the roads and the supply chain.
If it was humans who occasionally missed welds because they were hung over or were distracted because they had just been dumped by their husband, many would probably make it through inspection and end up on the road never to be noticed until an accident years later causes serious injury or death to the passenger and, for some reason, investigators actually dig into it and discover missing welds. Then, every car of the model/vintage remaining on the roads needs to be inspected for missing welds - which would probably be more expensive and reputation busting, even if not another single missing weld was found, than junking these 293 cars.
One of the advantages of automation is that it tends to make the same mistake over and over - humans are more random about their mistakes.
(I do wonder if they pull some usable components out of them - drivetrain, wheels, ECUs etc before crushing them. Probably it's not worth it as they would have to pay for removal and storage and introduce them into the supply chain for warranty repairs or similar - the supply chain is probably too inflexible to make that work. although, maybe they could use them in mechanic training...)
It is not mistreating someone to call them out for a serious error. A high contrast response is sometimes necessary to get their (and other's) attention and to communicate the severity of the infraction. The example I gave wasn't just "because Linus can" - it was a relevant message communicated at a fairly appropriate volume. Myself, I wouldn't have sent the message quite that way but that doesn't mean it was wrong. Everyone on the project at that point was well aware of Linus' personality and bluntness and if they didn't want to deal with such feedback when they did something really stupid and doubled down on it, they should have found something else to do. Over my career I've run across a number of people I would refuse to work "for" even if it meant I had to quit to accomplish that.
Sometimes when someone who shouldn't screw up does so with blatant disregard for the priorities of the project, it's useful to flame them to remind other people NOT to do the same thing.
One example is from about six years ago when Linus reminded everyone very crisply that one doesn't change userspace APIs willy nilly and then blame the applications that were broken by the change. I'm pretty sure that his response reinforced in many developers' minds that this was simply unacceptable and reminded them far more effectively than an unemotional purely technical observation would have.
'Grandfathered' plans are exempt from the 80% Medical Loss Ratio rule as are those with fewer than 1000 enrollees in a particular state or market. I believe all other conventional health insurance plans are subject to the restriction.
However, unless your company fails to meet the 80% MLR rule, most people wouldn't know about the MLR. The fact that your rates may have gone up may just reflect that either the insurer was paying out over the 80% and moved closer to paying out only 80% by raising rates and/or what they paid for medical services went up (perhaps because of acquiring a bunch of policy holders that had serious preexisting conditions or just because of increasing costs). They can also count the cost of "quality improvement activities" from the 80% - maybe your insurer is spending big on improving quality:)
In some markets, health Insurance companies have little reason to do what you describe. Under the PPACA they must pay out 80% of the incoming claims dollars for medical services. The other 20% covers all administrative costs (claims processing, salaries, utilities, negotiating networks, leases, computers, training, HR, claim review, approval review, etc) related to servicing existing customers and advertising (necessary to replace existing customers who pick a different plan, leave the service area, or die) and profits. If they pay out less than 80% premiums to medical claims, they have to rebate the difference back to the policy holders.
In fact, this limitation adds a perverse incentive to pay out MORE in claims, not less. If, hypothetically, an insurer doubles their premiums and doubles their medical pay outs (by not checking claims carefully, by approving questionable procedures, by not negotiating as hard as they could with medical providers) to keep within the 80% limit, the 20% they get to keep also doubles. Also, their expenses related to claims review and network negotiation probably drops as they they would just have to pay rebates back to policy holders if they deny claims and miss the 80% requirement. So, much - maybe more than all - of the increase in the 20% can go to pure profit.
Of course, if they raise their rates (and pay out more in claims), they may lose customers in a competitive market -- but in some areas there isn't a competitive market. For example, there are 1565 counties where there is only one provider on the exchange. In 2016, 30% of the participants in the Federal exchange had only ONE choice available to them. Also, since many people's premiums on the exchanges are highly subsidized by tax dollars, many people are not nearly as price sensitive as people who are paying their own way so raising rates in "one insurer" markets won't drive away business like it would in a conventional free market.
They are offering you unlimited data in the sense that you don't pay a different amount for data based on usage.
What they don't guarantee you is bandwidth -- just as they don't guarantee bandwidth on any plan including a 1GB/month plan.
Do you take unlimited to mean "Infinite" since that's what it means in the extreme? Even bandwidth of 100PB/sec 7/24 for a whole month doesn't give you that because 2.7e20 bytes << infinity bytes.
I'm not a big fan of the term 'unlimited' either, but now that everyone understands what it probably means, it really doesn't matter much. Would you prefer "Unmetered Data"? Perhaps "Fixed Price Data".
Obviously it is impossible, for a price that anyone other than a Saudi prince can afford, to provide mobile data at high speeds that is truly unlimited AND delivers full speed at all times. That's just the nature of the beast. In fact, it's effectively impossible to deliver the maximum mobile data rate everywhere due to signal strength issues even if bob4u2c is the ONLY user within 1000 miles.
What they mean by "unlimited" is you don't pay more for more data. As long as they only throttle when needed to avoid congestion and perhaps first to high usage users, this seems quite reasonable.
Nowhere in mobile plans do they guarantee bandwidth (as an OC-255 might for example) for the first byte or any other byte.
At this point, every person with at least two brain cells knows that 'unlimited' in mobile plans rarely means "full maximum bandwidth guaranteed 7x24 even if users actually use that bandwidth 7x24". People with only one or zero brain cells can just read the footnote that explains this.
I would acknowledge though that they should provide more details on the throttling and provide statistical information (perhaps in the form of maps as to what percentage of time at various times of day each area has been throttled over the prior month or so).
Maybe something like FirstNet which is clearly pointed out? (Except for, perhaps, the "low price" feature.)
Although, since these plans include "Preemption - priority access to the domestic AT&T 4G LTE network", they may not be able to offer them in California if it passes the pending "net neutrality" legislation (or any other state that legislates "net neutrality").
I don't see any reason for such plans to necessarily be "low price". Private enterprises should not be expected to provide governments with "bargains" that they wouldn't offer to other similar volume users. Should Ford sell SUVs to police departments cheaper than they would to a similar volume non-first responder, non-governmental organization? Why?
Actually, I read it elsewhere over a day ago in traditional media (and not from the source quoted in this story). So,/. isn't the only place to cover this. Sorry.
I suspect the board may demand that he effectively do that. Although probably not deleted -- just control relinquished. Perhaps he doesn't have the password and needs to forward things to tweet to a few people with attention spans, civility, and social filters more appropriate to an executive of a sizeable company than to a two year old in nursery school. If one of these these trusted people from every one of his ventures (SpaceX, Tesla, Boring, ???) approve, it gets sent it under Musk's account.
I find it interesting how Pedo Musk (if he can attribute that to someone without a shred of evidence, so can I -- and the fact that he thought to do suggests to me that maybe "pedo" is on his mind a lot, perhaps because he's worried if his laptop disks are properly encrypted?) worded part of this:
I knew the process of going private would be challenging, but it's clear that it would be even more time-consuming and distracting than initially anticipated.
By saying "I knew" he's taking credit for the fact that he was so smart as to know it would be challenging (which was actually obvious to analysts who published articles within an hour of his tweet). But when it comes to discovering that it was more challenging than he anticipated, he disowns that (it just now "clear").
He won't take responsibility here (as he rarely does). If he took responsibility for his FU, he would have shown ownership by writing:
I knew the process of going private would be challenging, but it's clear that it would be even more time-consuming and distracting than I initially anticipated.
It appears he had not discussed the concept with the board in any detail or as anymore than an abstract thought before his surprise announcement - so it was HE who failed to correctly anticipate the difficulty, not some amorphous unknown entity.
Agreed - but it is possible to believe there is a MoviePass model that works by selling the data about what movies you watch, or selling highly targeted ads (esp. to movies) based on what you watch and generating revenue from that.
One problem with the highly targeted movie ads revenue stream is if the ads are successful you will go see a movie you wouldn't have otherwise. Now if this replaced another movie you would have seen, one studio/producer gets one more movie view that another studio/producer lost and your decision is revenue neutral to MP (unless you would have originally gone to a cheap movie theater to see a second run movie and instead went to a full price theater to see the advertised "blockbuster") but they get to keep the ad revenue and that studio/producer (or others) will continue to buy ads due to their success. However if you see the movie in addition to other movies, MP pays for the movie and there's no possibility in a sustainable model that MP would routinely get more for the ads than what they pay for movie.
Another source of anticipated revenue could be the 'gym membership' phenomena -- you sign up and (maybe) use it for a short time and then stop going but just let auto debit continue. However, gym memberships have a sticky attribute in this case -- cancelling makes the person admit that they really are NOT going to start working out which is something they obviously thought they should do. People tend to be somewhat resistant to openly admitting (even to themselves) that they are unlikely to do what they know they should be doing and cancelling makes the person take that step. I don't think MoviePass has this level of stickiness -- the most you have to do is acknowledge that you thought you would see more movies if they were "free", but it turns out you just don't have the time to do so or discovered there aren't enough movies you want to see to make MP attractive -- that's an analytical decision that doesn't cast "moral" aspersions back on yourself.
"...it's too bad few of us use many data structures any more."
If someone writing a program doesn't think they use data structures, perhaps they should have gotten a CS degree from a decent university. Well, arguably, maybe if all they write are simple Hello World programs they are not "using" data structures explicitly (ignoring, of course, that they very stack that is used in the calling of printf() is a data structure).
I'd be interested in seeing a meaningful system where many programs in it don't use arrays, vectors, lists, queues, stacks, sets, etc.
Do most programmers still code up their own data structures outside of low level "systems" programming or real time systems, no. But they certainly "use" them and need to know the performance implications of which, for example, container class they pick.
Per Bloomberg, the the terms of the loan are pretty aggressive.
The lender can demand more than $3M be repaid on August 1 and the remainder on August 5. Also, MoviePass has a planned stock sale and proceeds from that must be used to repay the loan. If MoviePass is 48 hours late in paying, the debt will increase to 130%. If they pay late, they will pay a 15% annualized late fee as well.
This is not close to as restrictive as "non-compete" clauses in employee/employer agreements. These clauses (in the cases they are enforceable) typically restrict the employee's right to work for any competitor for some period of time. If enforced, this can make a specialist almost unemployable in some cases. As well, the action will be taken by the ex-employer against the employee -- resulting in the employee incurring legal expenses and, perhaps, requiring her to pay the ex-employer's legal fees as well if she loses the battle.
The agreement in question here is between franchise owners from the same franchisor. As such, it only potentially restricts an employee at one Burger King franchise from working for another Burger King franchise -- that employee is free to work at the other 95%+ of the restaurant jobs in the area that are NOT Burger King franchises. It's an agreement between the franchisees, not the employee and a franchise. As a result, any resulting legal action will not cost the employee legal fees or result in here having to pay the legal fees of the franchisee bringing the legal action in case of a violation.
If one entity owns many franchises (which is not unusual - in fact, one entity may own all the franchises for a particular brand within many miles), the owner is free to have a policy that no store manager is to hire an employee from another one of the owner's franchises unless the applicant has worked at their current store for at least two years. I don't think anyone would argue with that just as they would not argue that a large engineering company can't have a rule that workers can't transfer between groups at their own initiation more than once every two years (such a policy may be unwise, but it certainly is not illegal). If you consider that the franchisor is the owner of the brand, it controls many aspects of the franchisees operations (including cleanliness, consistency of food product, product offerings, employee training, and employee policies) and this is just one more such restriction on employee recruitment so franchise owners don't get screwed by other franchise owners (just as if one franchise was putting out horrible food, it would screw other franchises).
What about a common case where two businesses enter into a relationship for one to sell consulting or contract services to the other? It's very common for the contract to include terms that prohibit the purchaser of these services from poaching the employees of the supplier of the services without the approval of the supplier. Surely no one would argue that's not acceptable. This agreement, like the fast food case, is an agreement between two businesses, not a business and an employee, and doesn't prevent the ex-employee from doing anything or require them to do anything. Both the consultant and the fast food worker are free to apply for jobs at the client of the contracting firm or fast food franchise respectively and accept a job if offered -- they probably won't get hired because the potential employer will probably honor their contract with the other business, but if the potential employer is willing to fight a legal battle, they might make an offer. Note that in small markets and in specialty fields, the contractor/client restriction may narrow an employee's options way more than the fast food restriction does. There may be only three companies within 100 miles that need aeronautical engineers on permanent staff and if an aeronautical engineer ends up working for a job shop that places them into one of these three companies, 1/3 of the employee's future employment options (at least in the near term) have just vaporized. It's quite rare that 1/3 of the fast food jobs in an area with at least three fast food outlets would be at franchises of the same brand.
This does not seem to be nearly as insidious or restrictive as some are claiming.
While it would reduce labor for many repairs, it would likely increase initial manufacturing labor costs and increase the number of SKUs to stock for spares (which increases costs for both the manufacturer and dealers). Large integrated harnesses can be built by (the cheapest?) suppliers and be installed "on the line" more quickly than a bunch of discrete wiring.
The manufacturer cares much more about initial manufacturing cost than later repair costs - by reducing the former they can either keep the difference (more profit per car) and/or sell more cars because their pricing is more competitive (therefore increasing volume).
While this decision likely increases the cost of insurance slightly, consumers don't look that closely at that aspect and it's quite possible that if the manufacturer passes on a portion of their manufacturing cost savings that will more than compensate for the increased labor costs of replacing a complicated wiring harness. Most cars (well, before soy based insulation at least) never have any wiring harnesses replaced and many cars that would need that done would have been totaled even if the parts and labor for replacing the harness were free (for example after a fire or flood).
And you are missing the point of autopilots -- they LET you safely virtually ignore your course of travel for significant periods of time - at least tens of seconds. In fact, that's pretty much their reason for existing. That basic core function of autopilots is NOT safe with Tesla's so called "autopilot" so it's dishonest and dangerously misleading to call it "autopilot".
The difference here is the environment. In the case of open air or open water there are large distances between the thing under control of autopilot and fixed and moving things around it. That, of course, is not true when autos are driving on roads - autos in such environments are rarely more than a few seconds from a dangerous encounter with fixed or moving objects. This is especially true given that Tesla's autopilot seems to have a hard time figuring out where the road is and where a concrete wall is and is happy to unexpectedly steer the car into a solid object requiring the driver to respond to a bizarre and unexpected autonomous behavior that drivers of ordinary cars don't have to deal with. Blaming this product defect on the driver and Caltrans rather than accepting responsibility is right in line with the Musk playbook -- right up there with refusing to answer legitimate questions from analysts "because they are boring". Musk is seeming more and more unhinged -- the pressure and the realization that his baby is in dire trouble is probably getting to him.
I assume, from your arguments, that you would be okay with Tesla having named the feature "Hands Off Self Driving" -- and then warned drivers not to take their hands off the wheel or rely on the feature for much of anything. After all, the name just doesn't matter.
It's good that Tesla is finally trying to improve the safety of their bad design. It is unfortunate that they are so inexperienced and shallow in talent that this situation ever happened. They need better engineers and management and probably need to hire more human behavior experts (who have known for a very long time that things like level 3 automation where an operator must continuously monitor to detect the need to take over and do so almost instantly are dangerous due to how the human mind works).
Yes, someone would buy the scraps of Tesla in bankruptcy, but that's what usually happens in bankruptcy.
"Deep pocketed buyers" are "deep pocketed" because they don't lose money, they make it. They don't buy entities that are hemorrhaging unless they think they can turn around the company, not just because the tech is "cool". Tesla can only run in the red for so long, and soon there will be a lot of electric cars to choose from that have established dealer networks, parts supply chains, service networks, and massive manufacturing capabilities and expertise. Tesla will have to compete on price in the midrange market and that's a tough business for a newcomer.
If Tesla goes down the path I think is most likely, I don't expect the Tesla brand to be worth much and doubt that an established player would acquire the brand and continue to produce cars under that badge. Obviously Musk would no longer be associated with the Tesla name at that point and people are buying Musk's image, not Tesla's when they opt to buy a Tesla vehicle.
This is not the Apple model where there are a lot of fanboys who can afford to overpay, thereby yielding enormous profit margins to Apple, for their phones because they buying them as a fashion statement. There are a limited number of Tesla fanboys who can afford the large upfront premium Tesla hopes to get -- probably not enough to actually sell enough of overpriced Model 3s.
Customer satisfaction numbers don't mean all that much in this case. People who have paid for the "cool new thing" tend to be happy and overlook problems of all types -- it's rather like people whose babies are objectively simply ugly rarely recognize that fact and think their baby is the cutest baby they have ever seen. It's confirmation bias hard at work. Unfortunately for Tesla, they need to rely on an expanding their base to people who just want reliable, energy efficient, cost effective to and from work and, when they discover the high insurance costs for Teslas won't just overlook that or when they discover that service is very expensive because there independent mechanics are not nearly as available (in part because Tesla clamps down on the supply of parts and makes it hard for third parties to get some parts).
I hope that Tesla succeeds, I just think its quite likely they won't do so in the passenger car market. They may have a shot at an important, but limited, market in trucks or something else, but since that enterprise would have all the debt that Tesla has, that probably won't work out except by a chapter 11 bankruptcy where creditors get a buzzcut - and Musk certainly would almost certainly not have control of the resulting company.
Level 3 automation in its pure form is probably dead. Some established car companies (Ford I believe is one) has said they will not sell level 3 cars and will jump right to level 4 -- for exactly the reason that level 3 is inherently unsafe.
Suppose you manufacture shoddy automotive jackstands and put warning labels on them saying "Do not use in any situation where failure of jackstand would cause human injury" and a purchaser is fixing their brakes and is injured or killed due to your jackstand failing in ordinary use within its rated load capacity. Your warning is not likely to protect you from liability -- because simply by selling an automotive jackstand, it's understood by most people to have certain attributes -- including being able to safely get close to a car that is being held up by them -- that's why people buy them. This is similar to the concept of "implied warranty of merchantability" in the warranty world -- all products must be fit for their ordinary purpose.
The term "autopilot" is implying something Tesla is not selling and that is a big part of the problem here -- ordinary consumers, in spite of what appear to be CYA warnings (which cover everything under the sun) who have paid for "autopilot" reasonably expect it to be perform similarly to an "autopilot".
Worse, selling a product that you can easily make much safer with a software update and you fail to do can expose you to liability in spite of warnings. The fact that in the recent "smash into back of fire department maintenance truck" case, Tesla says "the driver had her hands off the wheel for 80 seconds before the crash" and that we now know that the "autopilot" actually increased speed while the software knew the driver was not using the "autopilot" in a safe way shows negligent autopilot design on the part of Tesla.
Tesla could easily make the "hands off wheel" warnings so annoying that they are virtually impossible to ignore and if the driver does so, after a few seconds, reduce speed slowly (perhaps turning on the brake lights to warn other cars) and notify the driver they must take control. Once the car has invoked the "I give up because you're not following the rules" mode the second time, autopilot should be disabled either for an extended time (weeks) and/or until the owner has gone to Tesla and gotten the necessary refresher course and signed off again on the restrictions. As well (and I don't know if it is) the feature should be protected so that only authorized drivers who have had the training can use it (perhaps by having userids and some sort of user verification) - and authorized autopilot drivers agree not to enable the feature for any other person (shifting liability to the person who does so by giving out their id/verification code or by using their biometric verification when someone else is driving the vehicle).
I tend to side with the manufacturer on many safety issues "caused" by misuse of a product. But, on this one, I can't bring myself to do so as the feature is so dangerous due to negligent design without reasonable safety checks and is named in a way that misleads consumers.
(Luckily for Tesla, I would probably be dismissed "for cause" from any jury that would be hearing such a case!).
The plaintiffs lawyers were probably acutely aware that if they didn't settle, the case might not have even gone to trial until after Tesla is bankrupt. Better $5M today than $0 later after having spent yet more money on the case.
Fortunately, Tesla's autocrash feature seems to have an attraction to large solid objects so others don't seem to be at much risk. Now, if in a future revision, it starts hunting Smart cars or MINI's as well, then "innocents" might suffer more injuries.
You are missing the key point of "autopilot" and the reason it's on boats and planes.
In the case of boats and planes, autopilots, regardless of how "sophisticated" they are, share one attribute -- they allow you to safely take your hands off the controls for significant periods of time (tens of seconds at least) and divert much/most of your attention to other matters (like looking at a chart). This will be preceded by the pilot making some sort of scan of the environment for hazards both fixed and mobile (in particular other boats and planes) by visual identification, radar, charts, etc and/or knowing that rules of the "road" (ATC imposed for planes) will insure a clear route.
Tesla autopilot fails to deliver on this expectation in two ways. First, the environment it is in coupled with its limited capabilities make it impossible to scan the environment in advance for hazards that will be encountered and that Teslapilot can't deal with (which, itself, appears quite difficult to predict). Second, it tends to run into stationary objects (fire engines, fire department maintenance trucks) and even, it appears, sometimes steers the car into them (gore points). If "autopilot" doesn't let you divert any attention from the road and, actually, makes you pay extra attention in case the car decides to steer into a fixed object, it simply is NOT "autopilot" as the typical consumer would expect it to be.
Slashdot Mirror
https://start.duckduckgo.com omits the DuckDuckGo self promotion etc.
I've not read the ruling, but from the summary it sounds like their hands were tied by a recent Supreme Court ruling. Give them a bit of time, they will figure out a way around that ruling and eventually decide such contracts are not lawful.
However in this case they knew that exactly 293 cars were affected and which ones they were. This made it easy to track down and remove the unsafe cars from the roads and the supply chain.
If it was humans who occasionally missed welds because they were hung over or were distracted because they had just been dumped by their husband, many would probably make it through inspection and end up on the road never to be noticed until an accident years later causes serious injury or death to the passenger and, for some reason, investigators actually dig into it and discover missing welds. Then, every car of the model/vintage remaining on the roads needs to be inspected for missing welds - which would probably be more expensive and reputation busting, even if not another single missing weld was found, than junking these 293 cars.
One of the advantages of automation is that it tends to make the same mistake over and over - humans are more random about their mistakes.
(I do wonder if they pull some usable components out of them - drivetrain, wheels, ECUs etc before crushing them. Probably it's not worth it as they would have to pay for removal and storage and introduce them into the supply chain for warranty repairs or similar - the supply chain is probably too inflexible to make that work. although, maybe they could use them in mechanic training...)
It is not mistreating someone to call them out for a serious error. A high contrast response is sometimes necessary to get their (and other's) attention and to communicate the severity of the infraction. The example I gave wasn't just "because Linus can" - it was a relevant message communicated at a fairly appropriate volume. Myself, I wouldn't have sent the message quite that way but that doesn't mean it was wrong. Everyone on the project at that point was well aware of Linus' personality and bluntness and if they didn't want to deal with such feedback when they did something really stupid and doubled down on it, they should have found something else to do. Over my career I've run across a number of people I would refuse to work "for" even if it meant I had to quit to accomplish that.
Sometimes when someone who shouldn't screw up does so with blatant disregard for the priorities of the project, it's useful to flame them to remind other people NOT to do the same thing.
One example is from about six years ago when Linus reminded everyone very crisply that one doesn't change userspace APIs willy nilly and then blame the applications that were broken by the change. I'm pretty sure that his response reinforced in many developers' minds that this was simply unacceptable and reminded them far more effectively than an unemotional purely technical observation would have.
'Grandfathered' plans are exempt from the 80% Medical Loss Ratio rule as are those with fewer than 1000 enrollees in a particular state or market. I believe all other conventional health insurance plans are subject to the restriction.
However, unless your company fails to meet the 80% MLR rule, most people wouldn't know about the MLR. The fact that your rates may have gone up may just reflect that either the insurer was paying out over the 80% and moved closer to paying out only 80% by raising rates and/or what they paid for medical services went up (perhaps because of acquiring a bunch of policy holders that had serious preexisting conditions or just because of increasing costs). They can also count the cost of "quality improvement activities" from the 80% - maybe your insurer is spending big on improving quality :)
In some markets, health Insurance companies have little reason to do what you describe. Under the PPACA they must pay out 80% of the incoming claims dollars for medical services. The other 20% covers all administrative costs (claims processing, salaries, utilities, negotiating networks, leases, computers, training, HR, claim review, approval review, etc) related to servicing existing customers and advertising (necessary to replace existing customers who pick a different plan, leave the service area, or die) and profits. If they pay out less than 80% premiums to medical claims, they have to rebate the difference back to the policy holders.
In fact, this limitation adds a perverse incentive to pay out MORE in claims, not less. If, hypothetically, an insurer doubles their premiums and doubles their medical pay outs (by not checking claims carefully, by approving questionable procedures, by not negotiating as hard as they could with medical providers) to keep within the 80% limit, the 20% they get to keep also doubles. Also, their expenses related to claims review and network negotiation probably drops as they they would just have to pay rebates back to policy holders if they deny claims and miss the 80% requirement. So, much - maybe more than all - of the increase in the 20% can go to pure profit.
Of course, if they raise their rates (and pay out more in claims), they may lose customers in a competitive market -- but in some areas there isn't a competitive market. For example, there are 1565 counties where there is only one provider on the exchange. In 2016, 30% of the participants in the Federal exchange had only ONE choice available to them. Also, since many people's premiums on the exchanges are highly subsidized by tax dollars, many people are not nearly as price sensitive as people who are paying their own way so raising rates in "one insurer" markets won't drive away business like it would in a conventional free market.
Okay, so then they just shut you off (i.e., throttle to 0mbs) after that threshold is reached since that's all they advertised.
They are offering you unlimited data in the sense that you don't pay a different amount for data based on usage.
What they don't guarantee you is bandwidth -- just as they don't guarantee bandwidth on any plan including a 1GB/month plan.
Do you take unlimited to mean "Infinite" since that's what it means in the extreme? Even bandwidth of 100PB/sec 7/24 for a whole month doesn't give you that because 2.7e20 bytes << infinity bytes.
I'm not a big fan of the term 'unlimited' either, but now that everyone understands what it probably means, it really doesn't matter much. Would you prefer "Unmetered Data"? Perhaps "Fixed Price Data".
Obviously it is impossible, for a price that anyone other than a Saudi prince can afford, to provide mobile data at high speeds that is truly unlimited AND delivers full speed at all times. That's just the nature of the beast. In fact, it's effectively impossible to deliver the maximum mobile data rate everywhere due to signal strength issues even if bob4u2c is the ONLY user within 1000 miles.
What they mean by "unlimited" is you don't pay more for more data. As long as they only throttle when needed to avoid congestion and perhaps first to high usage users, this seems quite reasonable.
Nowhere in mobile plans do they guarantee bandwidth (as an OC-255 might for example) for the first byte or any other byte.
At this point, every person with at least two brain cells knows that 'unlimited' in mobile plans rarely means "full maximum bandwidth guaranteed 7x24 even if users actually use that bandwidth 7x24". People with only one or zero brain cells can just read the footnote that explains this.
I would acknowledge though that they should provide more details on the throttling and provide statistical information (perhaps in the form of maps as to what percentage of time at various times of day each area has been throttled over the prior month or so).
Maybe something like FirstNet which is clearly pointed out? (Except for, perhaps, the "low price" feature.)
Although, since these plans include "Preemption - priority access to the domestic AT&T 4G LTE network", they may not be able to offer them in California if it passes the pending "net neutrality" legislation (or any other state that legislates "net neutrality").
I don't see any reason for such plans to necessarily be "low price". Private enterprises should not be expected to provide governments with "bargains" that they wouldn't offer to other similar volume users. Should Ford sell SUVs to police departments cheaper than they would to a similar volume non-first responder, non-governmental organization? Why?
Actually, I read it elsewhere over a day ago in traditional media (and not from the source quoted in this story). So, /. isn't the only place to cover this. Sorry.
I suspect the board may demand that he effectively do that. Although probably not deleted -- just control relinquished. Perhaps he doesn't have the password and needs to forward things to tweet to a few people with attention spans, civility, and social filters more appropriate to an executive of a sizeable company than to a two year old in nursery school. If one of these these trusted people from every one of his ventures (SpaceX, Tesla, Boring, ???) approve, it gets sent it under Musk's account.
I find it interesting how Pedo Musk (if he can attribute that to someone without a shred of evidence, so can I -- and the fact that he thought to do suggests to me that maybe "pedo" is on his mind a lot, perhaps because he's worried if his laptop disks are properly encrypted?) worded part of this:
By saying "I knew" he's taking credit for the fact that he was so smart as to know it would be challenging (which was actually obvious to analysts who published articles within an hour of his tweet). But when it comes to discovering that it was more challenging than he anticipated, he disowns that (it just now "clear").
He won't take responsibility here (as he rarely does). If he took responsibility for his FU, he would have shown ownership by writing:
It appears he had not discussed the concept with the board in any detail or as anymore than an abstract thought before his surprise announcement - so it was HE who failed to correctly anticipate the difficulty, not some amorphous unknown entity.
Agreed - but it is possible to believe there is a MoviePass model that works by selling the data about what movies you watch, or selling highly targeted ads (esp. to movies) based on what you watch and generating revenue from that.
One problem with the highly targeted movie ads revenue stream is if the ads are successful you will go see a movie you wouldn't have otherwise. Now if this replaced another movie you would have seen, one studio/producer gets one more movie view that another studio/producer lost and your decision is revenue neutral to MP (unless you would have originally gone to a cheap movie theater to see a second run movie and instead went to a full price theater to see the advertised "blockbuster") but they get to keep the ad revenue and that studio/producer (or others) will continue to buy ads due to their success. However if you see the movie in addition to other movies, MP pays for the movie and there's no possibility in a sustainable model that MP would routinely get more for the ads than what they pay for movie.
Another source of anticipated revenue could be the 'gym membership' phenomena -- you sign up and (maybe) use it for a short time and then stop going but just let auto debit continue. However, gym memberships have a sticky attribute in this case -- cancelling makes the person admit that they really are NOT going to start working out which is something they obviously thought they should do. People tend to be somewhat resistant to openly admitting (even to themselves) that they are unlikely to do what they know they should be doing and cancelling makes the person take that step. I don't think MoviePass has this level of stickiness -- the most you have to do is acknowledge that you thought you would see more movies if they were "free", but it turns out you just don't have the time to do so or discovered there aren't enough movies you want to see to make MP attractive -- that's an analytical decision that doesn't cast "moral" aspersions back on yourself.
If someone writing a program doesn't think they use data structures, perhaps they should have gotten a CS degree from a decent university. Well, arguably, maybe if all they write are simple Hello World programs they are not "using" data structures explicitly (ignoring, of course, that they very stack that is used in the calling of printf() is a data structure).
I'd be interested in seeing a meaningful system where many programs in it don't use arrays, vectors, lists, queues, stacks, sets, etc.
Do most programmers still code up their own data structures outside of low level "systems" programming or real time systems, no. But they certainly "use" them and need to know the performance implications of which, for example, container class they pick.
Per Bloomberg, the the terms of the loan are pretty aggressive.
The lender can demand more than $3M be repaid on August 1 and the remainder on August 5. Also, MoviePass has a planned stock sale and proceeds from that must be used to repay the loan. If MoviePass is 48 hours late in paying, the debt will increase to 130%. If they pay late, they will pay a 15% annualized late fee as well.
This is not close to as restrictive as "non-compete" clauses in employee/employer agreements. These clauses (in the cases they are enforceable) typically restrict the employee's right to work for any competitor for some period of time. If enforced, this can make a specialist almost unemployable in some cases. As well, the action will be taken by the ex-employer against the employee -- resulting in the employee incurring legal expenses and, perhaps, requiring her to pay the ex-employer's legal fees as well if she loses the battle.
The agreement in question here is between franchise owners from the same franchisor. As such, it only potentially restricts an employee at one Burger King franchise from working for another Burger King franchise -- that employee is free to work at the other 95%+ of the restaurant jobs in the area that are NOT Burger King franchises. It's an agreement between the franchisees, not the employee and a franchise. As a result, any resulting legal action will not cost the employee legal fees or result in here having to pay the legal fees of the franchisee bringing the legal action in case of a violation.
If one entity owns many franchises (which is not unusual - in fact, one entity may own all the franchises for a particular brand within many miles), the owner is free to have a policy that no store manager is to hire an employee from another one of the owner's franchises unless the applicant has worked at their current store for at least two years. I don't think anyone would argue with that just as they would not argue that a large engineering company can't have a rule that workers can't transfer between groups at their own initiation more than once every two years (such a policy may be unwise, but it certainly is not illegal). If you consider that the franchisor is the owner of the brand, it controls many aspects of the franchisees operations (including cleanliness, consistency of food product, product offerings, employee training, and employee policies) and this is just one more such restriction on employee recruitment so franchise owners don't get screwed by other franchise owners (just as if one franchise was putting out horrible food, it would screw other franchises).
What about a common case where two businesses enter into a relationship for one to sell consulting or contract services to the other? It's very common for the contract to include terms that prohibit the purchaser of these services from poaching the employees of the supplier of the services without the approval of the supplier. Surely no one would argue that's not acceptable. This agreement, like the fast food case, is an agreement between two businesses, not a business and an employee, and doesn't prevent the ex-employee from doing anything or require them to do anything. Both the consultant and the fast food worker are free to apply for jobs at the client of the contracting firm or fast food franchise respectively and accept a job if offered -- they probably won't get hired because the potential employer will probably honor their contract with the other business, but if the potential employer is willing to fight a legal battle, they might make an offer. Note that in small markets and in specialty fields, the contractor/client restriction may narrow an employee's options way more than the fast food restriction does. There may be only three companies within 100 miles that need aeronautical engineers on permanent staff and if an aeronautical engineer ends up working for a job shop that places them into one of these three companies, 1/3 of the employee's future employment options (at least in the near term) have just vaporized. It's quite rare that 1/3 of the fast food jobs in an area with at least three fast food outlets would be at franchises of the same brand.
This does not seem to be nearly as insidious or restrictive as some are claiming.
While it would reduce labor for many repairs, it would likely increase initial manufacturing labor costs and increase the number of SKUs to stock for spares (which increases costs for both the manufacturer and dealers). Large integrated harnesses can be built by (the cheapest?) suppliers and be installed "on the line" more quickly than a bunch of discrete wiring.
The manufacturer cares much more about initial manufacturing cost than later repair costs - by reducing the former they can either keep the difference (more profit per car) and/or sell more cars because their pricing is more competitive (therefore increasing volume).
While this decision likely increases the cost of insurance slightly, consumers don't look that closely at that aspect and it's quite possible that if the manufacturer passes on a portion of their manufacturing cost savings that will more than compensate for the increased labor costs of replacing a complicated wiring harness. Most cars (well, before soy based insulation at least) never have any wiring harnesses replaced and many cars that would need that done would have been totaled even if the parts and labor for replacing the harness were free (for example after a fire or flood).
And you are missing the point of autopilots -- they LET you safely virtually ignore your course of travel for significant periods of time - at least tens of seconds. In fact, that's pretty much their reason for existing. That basic core function of autopilots is NOT safe with Tesla's so called "autopilot" so it's dishonest and dangerously misleading to call it "autopilot".
The difference here is the environment. In the case of open air or open water there are large distances between the thing under control of autopilot and fixed and moving things around it. That, of course, is not true when autos are driving on roads - autos in such environments are rarely more than a few seconds from a dangerous encounter with fixed or moving objects. This is especially true given that Tesla's autopilot seems to have a hard time figuring out where the road is and where a concrete wall is and is happy to unexpectedly steer the car into a solid object requiring the driver to respond to a bizarre and unexpected autonomous behavior that drivers of ordinary cars don't have to deal with. Blaming this product defect on the driver and Caltrans rather than accepting responsibility is right in line with the Musk playbook -- right up there with refusing to answer legitimate questions from analysts "because they are boring". Musk is seeming more and more unhinged -- the pressure and the realization that his baby is in dire trouble is probably getting to him.
I assume, from your arguments, that you would be okay with Tesla having named the feature "Hands Off Self Driving" -- and then warned drivers not to take their hands off the wheel or rely on the feature for much of anything. After all, the name just doesn't matter.
It's good that Tesla is finally trying to improve the safety of their bad design. It is unfortunate that they are so inexperienced and shallow in talent that this situation ever happened. They need better engineers and management and probably need to hire more human behavior experts (who have known for a very long time that things like level 3 automation where an operator must continuously monitor to detect the need to take over and do so almost instantly are dangerous due to how the human mind works).
Yes, someone would buy the scraps of Tesla in bankruptcy, but that's what usually happens in bankruptcy.
"Deep pocketed buyers" are "deep pocketed" because they don't lose money, they make it. They don't buy entities that are hemorrhaging unless they think they can turn around the company, not just because the tech is "cool". Tesla can only run in the red for so long, and soon there will be a lot of electric cars to choose from that have established dealer networks, parts supply chains, service networks, and massive manufacturing capabilities and expertise. Tesla will have to compete on price in the midrange market and that's a tough business for a newcomer.
If Tesla goes down the path I think is most likely, I don't expect the Tesla brand to be worth much and doubt that an established player would acquire the brand and continue to produce cars under that badge. Obviously Musk would no longer be associated with the Tesla name at that point and people are buying Musk's image, not Tesla's when they opt to buy a Tesla vehicle.
This is not the Apple model where there are a lot of fanboys who can afford to overpay, thereby yielding enormous profit margins to Apple, for their phones because they buying them as a fashion statement. There are a limited number of Tesla fanboys who can afford the large upfront premium Tesla hopes to get -- probably not enough to actually sell enough of overpriced Model 3s.
Customer satisfaction numbers don't mean all that much in this case. People who have paid for the "cool new thing" tend to be happy and overlook problems of all types -- it's rather like people whose babies are objectively simply ugly rarely recognize that fact and think their baby is the cutest baby they have ever seen. It's confirmation bias hard at work. Unfortunately for Tesla, they need to rely on an expanding their base to people who just want reliable, energy efficient, cost effective to and from work and, when they discover the high insurance costs for Teslas won't just overlook that or when they discover that service is very expensive because there independent mechanics are not nearly as available (in part because Tesla clamps down on the supply of parts and makes it hard for third parties to get some parts).
I hope that Tesla succeeds, I just think its quite likely they won't do so in the passenger car market. They may have a shot at an important, but limited, market in trucks or something else, but since that enterprise would have all the debt that Tesla has, that probably won't work out except by a chapter 11 bankruptcy where creditors get a buzzcut - and Musk certainly would almost certainly not have control of the resulting company.
Level 3 automation in its pure form is probably dead. Some established car companies (Ford I believe is one) has said they will not sell level 3 cars and will jump right to level 4 -- for exactly the reason that level 3 is inherently unsafe.
Suppose you manufacture shoddy automotive jackstands and put warning labels on them saying "Do not use in any situation where failure of jackstand would cause human injury" and a purchaser is fixing their brakes and is injured or killed due to your jackstand failing in ordinary use within its rated load capacity. Your warning is not likely to protect you from liability -- because simply by selling an automotive jackstand, it's understood by most people to have certain attributes -- including being able to safely get close to a car that is being held up by them -- that's why people buy them. This is similar to the concept of "implied warranty of merchantability" in the warranty world -- all products must be fit for their ordinary purpose.
The term "autopilot" is implying something Tesla is not selling and that is a big part of the problem here -- ordinary consumers, in spite of what appear to be CYA warnings (which cover everything under the sun) who have paid for "autopilot" reasonably expect it to be perform similarly to an "autopilot".
Worse, selling a product that you can easily make much safer with a software update and you fail to do can expose you to liability in spite of warnings. The fact that in the recent "smash into back of fire department maintenance truck" case, Tesla says "the driver had her hands off the wheel for 80 seconds before the crash" and that we now know that the "autopilot" actually increased speed while the software knew the driver was not using the "autopilot" in a safe way shows negligent autopilot design on the part of Tesla.
Tesla could easily make the "hands off wheel" warnings so annoying that they are virtually impossible to ignore and if the driver does so, after a few seconds, reduce speed slowly (perhaps turning on the brake lights to warn other cars) and notify the driver they must take control. Once the car has invoked the "I give up because you're not following the rules" mode the second time, autopilot should be disabled either for an extended time (weeks) and/or until the owner has gone to Tesla and gotten the necessary refresher course and signed off again on the restrictions. As well (and I don't know if it is) the feature should be protected so that only authorized drivers who have had the training can use it (perhaps by having userids and some sort of user verification) - and authorized autopilot drivers agree not to enable the feature for any other person (shifting liability to the person who does so by giving out their id/verification code or by using their biometric verification when someone else is driving the vehicle).
I tend to side with the manufacturer on many safety issues "caused" by misuse of a product. But, on this one, I can't bring myself to do so as the feature is so dangerous due to negligent design without reasonable safety checks and is named in a way that misleads consumers.
(Luckily for Tesla, I would probably be dismissed "for cause" from any jury that would be hearing such a case!).
The plaintiffs lawyers were probably acutely aware that if they didn't settle, the case might not have even gone to trial until after Tesla is bankrupt. Better $5M today than $0 later after having spent yet more money on the case.
Fortunately, Tesla's autocrash feature seems to have an attraction to large solid objects so others don't seem to be at much risk. Now, if in a future revision, it starts hunting Smart cars or MINI's as well, then "innocents" might suffer more injuries.
You are missing the key point of "autopilot" and the reason it's on boats and planes.
In the case of boats and planes, autopilots, regardless of how "sophisticated" they are, share one attribute -- they allow you to safely take your hands off the controls for significant periods of time (tens of seconds at least) and divert much/most of your attention to other matters (like looking at a chart). This will be preceded by the pilot making some sort of scan of the environment for hazards both fixed and mobile (in particular other boats and planes) by visual identification, radar, charts, etc and/or knowing that rules of the "road" (ATC imposed for planes) will insure a clear route.
Tesla autopilot fails to deliver on this expectation in two ways. First, the environment it is in coupled with its limited capabilities make it impossible to scan the environment in advance for hazards that will be encountered and that Teslapilot can't deal with (which, itself, appears quite difficult to predict). Second, it tends to run into stationary objects (fire engines, fire department maintenance trucks) and even, it appears, sometimes steers the car into them (gore points). If "autopilot" doesn't let you divert any attention from the road and, actually, makes you pay extra attention in case the car decides to steer into a fixed object, it simply is NOT "autopilot" as the typical consumer would expect it to be.