Back on August 9, I noted that the Atlanta Fed at their GDPNow site had given a preliminary forecast of third quarter GDP of 3.8% annualized. I referred to it as Grossly Distorted Prediction and noted I would be surprised if the GDP was ultimately half that for the quarter, i.e. 1.9%. Well, as of today, the Atlanta Fed is down to 2.4% and dropping fast. They will continue to post third quarter numbers through October, so I might make my guesstimate after all.
https://www.frbatlanta.org/-/media/Documents/cqer/researchcq/gdpnow/RealGDPTrackingSlides.pdf
October 3 update
Now down to 2.2% after a dismal construction spending report today.
October 7 update
2.1%
Friday, September 30, 2016
Tuesday, September 27, 2016
Gentelmen, Start Your . . . Lawsuits!
Well, it was just a matter of time. Tesla has been sued for a fatal crash in China allegedly involving its AutoPilot feature.
http://fortune.com/2016/09/15/tesla-autopilot-crash-china/
The lawsuit seeks very little monetarily and is reportedly designed primarily to just bring attention to the problems with the system. Tesla claims there is no way to know if the AutoPilot was engaged due to the car being too damaged in the crash. On the other hand, in the more recent fatal crash in the Netherlands, Tesla confirmed AutoPilot was not engaged and officials are not refuting it.
http://phys.org/news/2016-09-dutch-police-probe-fatal-tesla.html
Tesla agrees, however, that the fatal crash in Florida this year was with AutoPilot engaged, but asserts the driver was not using it properly. Moreover, it claims to have now issued a software upgrade that eliminates the issue that resulted in that crash.
I have reported before on the likelihood of lawsuits over this. Whether the China suit goes anywhere is anyone's guess. Either way, others will follow.
I have read some commentators' views that lawsuits are not that likely as the legal liability is not that easy or cheap to prove. Experts will be needed and will be expensive and state-of-the-art defenses may be available. All true, but never underestimate the drive of the plaintiffs' bar or its resources. Moreover, some expert will see the light and realize they can make a nice living testifying in these cases.
One might argue that different approaches being taken to autonomous cars provides plenty ammunition for an expert. For example, Google is designing Level 4 cars with no steering wheel and no chance for human intervention. It believes the interaction between a human driver and somewhat autonomous features can only lead to problems. And just maybe Tesla is proving that to be true.
Now I understand that Tesla has plenty of warnings a driver has to go through to even engage AutoPilot, that the driver still has to touch the wheel occasionally, that they are regularly doing upgrades, yada, yada, yada . . . All these yadas probably add up to a nice defense against the driver of the Tesla who is injured or killed. But what about the first time the Tesla plows into another car or people? It will eventually happen. Tesla will point the finger at the "driver" and the lawyers will point the finger at the deep pocket and the expert will say it is simply foolish to have the autonomous/human interaction as it creates too many variables. Poof, there it is. Mind you, these cars likely are safer than those with drivers in full control, but I still think Tesla is conducting a dangerous experiment. No doubt in doing so they are collecting enormous amounts of data with which to tweak their software, which is perhaps enough economic incentive for Tesla to take this chance. Only time will tell.
Let's conservatively assume the number of cars needed only goes down 50%. Well, now you have 50% less for parts suppliers to supply, for manufacturers to build, for insurers to insure, etc. Indeed, car retail will largely disappear as it will simply be large corporate fleets of autonomous vehicles serving the public, so say goodbye to your local car dealer. That is a whole lot of missing jobs and dollars - and a whole lot of empty retail parking space that is no longer needed.
Let's say you are Ford. Ford has indicated it is focusing on building autonomous cars for hire, not for private sale. This makes a good bit of sense as it leaves control and upkeep of these sophisticated machines in the hands of a few companies that know what they are doing. Now it is not clear to me if Ford is going to work with the likes of Uber or whether it may launch its own fleet. The latter makes sense to me.
The approach of selling these autonomous cars to individuals seems financially unwise. The cars may cost more and have a higher per vehicle profit but if you are making half as many, your profits go down. And liability - absent legislative intervention - is shifting largely to the manufacturer. So you have lower sales profits with which to pay what will likely be enormous premiums for liability insurance. Not per se a good business model.
So why not make your profit simply offering the cars you make for hire. People sign up with Ford and get charged per trip. Ford makes its money from this without ever selling a vehicle retail. It has fewer factories to maintain, fewer workers to pay and a regular income stream moving people around. Perhaps not the profits it is seeing from selling cars today, but it has to do something to fill the void that is coming.
And there will be voids for a lot of different businesses to fill if they can. Any wonder that those national car insurers are now emphasizing in adds their other services like loans and such? Perhaps they are seeing the writing on the wall. After all, studies predict car insurance premiums are going to go down 60% over the next 15-20 years and 80% over the next 25. We are talking about $200 billion in premiums, 80% of which might disappear in relatively short order. Ouch!
http://www.latimes.com/business/la-fi-agenda-driverless-insurance-20160620-snap-story.html
These are interesting times my friends. In an upcoming post I will endeavor to note some of the key benefits to autonomous cars beyond the safety benefits. For example, that two car garage might make a nice playroom for the kids.
http://fortune.com/2016/09/15/tesla-autopilot-crash-china/
The lawsuit seeks very little monetarily and is reportedly designed primarily to just bring attention to the problems with the system. Tesla claims there is no way to know if the AutoPilot was engaged due to the car being too damaged in the crash. On the other hand, in the more recent fatal crash in the Netherlands, Tesla confirmed AutoPilot was not engaged and officials are not refuting it.
http://phys.org/news/2016-09-dutch-police-probe-fatal-tesla.html
Tesla agrees, however, that the fatal crash in Florida this year was with AutoPilot engaged, but asserts the driver was not using it properly. Moreover, it claims to have now issued a software upgrade that eliminates the issue that resulted in that crash.
I have reported before on the likelihood of lawsuits over this. Whether the China suit goes anywhere is anyone's guess. Either way, others will follow.
I have read some commentators' views that lawsuits are not that likely as the legal liability is not that easy or cheap to prove. Experts will be needed and will be expensive and state-of-the-art defenses may be available. All true, but never underestimate the drive of the plaintiffs' bar or its resources. Moreover, some expert will see the light and realize they can make a nice living testifying in these cases.
One might argue that different approaches being taken to autonomous cars provides plenty ammunition for an expert. For example, Google is designing Level 4 cars with no steering wheel and no chance for human intervention. It believes the interaction between a human driver and somewhat autonomous features can only lead to problems. And just maybe Tesla is proving that to be true.
Now I understand that Tesla has plenty of warnings a driver has to go through to even engage AutoPilot, that the driver still has to touch the wheel occasionally, that they are regularly doing upgrades, yada, yada, yada . . . All these yadas probably add up to a nice defense against the driver of the Tesla who is injured or killed. But what about the first time the Tesla plows into another car or people? It will eventually happen. Tesla will point the finger at the "driver" and the lawyers will point the finger at the deep pocket and the expert will say it is simply foolish to have the autonomous/human interaction as it creates too many variables. Poof, there it is. Mind you, these cars likely are safer than those with drivers in full control, but I still think Tesla is conducting a dangerous experiment. No doubt in doing so they are collecting enormous amounts of data with which to tweak their software, which is perhaps enough economic incentive for Tesla to take this chance. Only time will tell.
Show Me The Money
Now autonomous cars are coming whether you like it or not and some entities hope to make a lot of mullah off of them. But there are several categories of companies where you have to scratch your head. Car manufacturers, for example. Sure, they will make these cars and sell them, but the obvious eventual plan here is that individuals will no longer own a car or at least no longer need to own one. They will simply summon an autonomous car from Uber or Ford or whoever when they need it, specifying why they need it so an appropriate vehicle shows up. They will then be taken to their destination and the car will proceed to its next pick-up. Instead of sitting idle 95% of the time, the car will be in use the majority of time, perhaps getting maintenance and such during the slow nighttime shifts. Overall, far fewer cars will be needed, perhaps only a third as many.Let's conservatively assume the number of cars needed only goes down 50%. Well, now you have 50% less for parts suppliers to supply, for manufacturers to build, for insurers to insure, etc. Indeed, car retail will largely disappear as it will simply be large corporate fleets of autonomous vehicles serving the public, so say goodbye to your local car dealer. That is a whole lot of missing jobs and dollars - and a whole lot of empty retail parking space that is no longer needed.
Let's say you are Ford. Ford has indicated it is focusing on building autonomous cars for hire, not for private sale. This makes a good bit of sense as it leaves control and upkeep of these sophisticated machines in the hands of a few companies that know what they are doing. Now it is not clear to me if Ford is going to work with the likes of Uber or whether it may launch its own fleet. The latter makes sense to me.
The approach of selling these autonomous cars to individuals seems financially unwise. The cars may cost more and have a higher per vehicle profit but if you are making half as many, your profits go down. And liability - absent legislative intervention - is shifting largely to the manufacturer. So you have lower sales profits with which to pay what will likely be enormous premiums for liability insurance. Not per se a good business model.
So why not make your profit simply offering the cars you make for hire. People sign up with Ford and get charged per trip. Ford makes its money from this without ever selling a vehicle retail. It has fewer factories to maintain, fewer workers to pay and a regular income stream moving people around. Perhaps not the profits it is seeing from selling cars today, but it has to do something to fill the void that is coming.
And there will be voids for a lot of different businesses to fill if they can. Any wonder that those national car insurers are now emphasizing in adds their other services like loans and such? Perhaps they are seeing the writing on the wall. After all, studies predict car insurance premiums are going to go down 60% over the next 15-20 years and 80% over the next 25. We are talking about $200 billion in premiums, 80% of which might disappear in relatively short order. Ouch!
http://www.latimes.com/business/la-fi-agenda-driverless-insurance-20160620-snap-story.html
These are interesting times my friends. In an upcoming post I will endeavor to note some of the key benefits to autonomous cars beyond the safety benefits. For example, that two car garage might make a nice playroom for the kids.
Subscribe to:
Posts (Atom)
