SOoo much to unpack from yesterday. Ok, lets take this one at a time. First the stock jump to over $1000 a share. In less than a week, TSLA went from $775 to close at $1080 today with a peak at $1094. This 25% jump is pretty crazy. So what is going on? Two things. Tesla made it numbers starting the run and then an unusual run on call options.

“Investors spent $5.11 billion on call options, compared with a total of $279 million on puts, according to an estimate by Nomura Securities. The firm calls the 18-to-1 ratio “unheard of.” And all the spike in calls means options dealers needed to snap up shares to avoid unwanted risk.”

The Motley Fool

Second Hertz placed an order for 100k cars!! Its this second event that is really interesting to me. So yes this is a great marketing move from a back-from-the-dead car rental company, but this also places a lot of pressure on other car rental services to match this offering. Does not necessary mean they will buy Tesla, but most likely that will part of the mix as other car manufactures try to get their EV exposed to the consumer market. As such, I would not be surprised to see other orders from other services in the near future. Does the mean the stock price will go up even more. No idea.

In other Tesla news, FSD started it first large release. Unfortunately, 10.2 had a pretty bad bug, ( I experienced it) and they had to pull it back. But soon after they released 10.3. So far so good! Yes its VERY much a beta so you have to pay very close attention. Big issue I noticed right off is the system does not always understand the concept of driveway slope paving when entering a community. It thinks it still on a road (according to Google Maps) and not only does not slow down, but unless you stop it, it can curb the tires pretty bad as its cuts the turn.

Other issue is when turning right. Often there is space for the car to occupy to bypass other cars in the right hand lane going straight. However, FSD does not understand that. That is an issue when cars behind you get confused that you turned on your blinker but did not get to the far right side of the lane.

Another odd one you will see in the video is when the car thought it was stuck trying to turn right and decided to go around. Only issue is, if you try that, you will not make your right hand turn anymore. I should have let it make the maneuver just to see what happens.

You can check out the video below. I cut most of the dead time out and only did a very short drive, but you can get a sense for its pretty quickly. Key note here. I have a sub 10k vin model 3. So its running on the older camera configuration (I think) so that may or may not play into performance.

So as far as I know, I may be the first person in history to use the new V10 Tesla Smart Summon on the Samsung Galaxy Fold phone. So shameless plug, I work at Samsung and I think the Fold is amazing. Yea it has its issues, its new, but I love it. I’ll write a post about it another day. So, that aside, I also did a few experiments with smart summon and got a mixed bag. So I shot a few vids and posted them. The tl:dr, GPS is not very accurate and if you are in an area with lots of building, it can through it off a few feet. If you are in a parking garage, its far worse. Next it sees many open spaces as an intersection and thus will pause. Better safe than sorry I guess. It also can ‘trap’ itself if it does not back out of the parking spot correctly. This has only happened once, but then again, I have only just started playing with it. That aside, it works most of the time. Sometimes you will have to restart your app because it will get ‘stuck’ after performing a summon.

First experiment went pretty well. The car moved from the parking space to me with only a few pauses. Did not make it all the way. I think it was not sure where I was given how odd the parking is here. I made it easy by backing the car in, which is how I park most of the time anyway. You can see on the large screen that the car is sending back what it ‘sees’, not video mind you. I would LOVE to get video instead, but may need a special encoder on the car for that to get the data load low enough. So test 1 was pretty good.

Test 2. As you can see on the Fold, this time I placed a pin and had the car come to the pin. This worked pretty well as well. But again I made it pretty easy since I backed the car in. Still, pretty cool! There is a distance limit, but hopefully that will get expanded as time goes on.

Test 3. I made this a little harder, and unfortunately, it failed this one. In this test I put it in a parking garage and parked next to a pillar. The pillar got it backing up incorrectly. Its also possible the GPS was messed up and it was thinking I was somewhere else, but either way, once it gets messed up as far as local path-finding, it has a hard time recovering. In this case it tried to leave the garage. I had to stop it since it was about to hit the pillar.

Test 4. In this test, I wanted to get away from the buildings and try to find a more open parking area so that GPS will not have reflection issues. That proved to be accurate, however, it saw each intersection in the parking area as perhaps normal intersections? Paused a few second each time. I do not mind once, but it seems to see far too many intersections. You can see this towards the end of the film.

Test 5, final test. I did this one because I read a comment on Reddit that made me want to get a film with our security robot. Two self driving systems living in a harmonious world… or something like that. Worked pretty well again, but this time, I was simply after a nice shot.

So there you go. Pretty impressive so far. The first experiment was actually last night in a different parking lot after shopping. Worked pretty much the same. It reversed out, and headed over. However, it tends to head straight for you and stop in the middle of lane. You need to be quick to get to the car. Do NOT recommend of its busy.

I remember standing in line to put my $1000 down on a car I have not even seen.  The excitement was like the first release of the iPhone, but without the wait.  So jump 2 years and my M3day was last month.  I love the car, although I wish the ride was smoother but I could fix that by changing the shocks… and I wont.  However, what I have found is that I love showing people the car.  I get excited talking about the fact that there is no dash.  LOVE showing the self driving.  The new electric motor the M3 uses, the smooth acceleration, how it looks, even the cool way the air vents operate.

Now there are things I would love to change now that I’m 30 days into my M3day.  I would really like to edit the UI.  Perhaps one day they will provide an API where I can code up my own via markup or directly in c++ or however they do it.  I’m also waiting for the self drive to improve and expand.  For example, the system should be smart enough to lean into a turn.  Right now the car still tried to keep the center line in a turn and ends up hugging the far line a bit too much for me and the unlucky person driving next to me.  This should be a simple algorithm fix.  (I work in computer vision for VR for a living BTW).  These details not withstanding, watching peoples eye light up as I let the car drive itself is so much fun.  I also love when people first sit in the car.  Its one thing to tell people, ‘no instrument panel, just a screen’ and then having people actually sit in it.   The comment I get all the time is, ‘ah so this is the future’.  And yes, I think it is.   Just like when I first got the iPhone and jailbroke it, I could see, this is where all phones are going to go.   The M3 is the first car in my mind that will signal, this is where we are going and I love sharing that.

BTW, why not say this about the S?  I drove the S a few times and really loved it, and it felt like really great electric car, but because of the dash, I still felt like I was in a normal car.  Also, the car felt too big, too many ‘features’ that were not really necessary, etc.  In other words, like the first iPhone.  Really cool, but not quite there yet. The dash-less experience with the M3, smaller more practical, only things on the car is what is needed on the car felt like a much cleaner design and more inline to what I would want in a car.  So like my iPhone 3G, yes the 2G gave me the first glips of the future, but the 3G is where I really experienced it, and so with the M3.

From Reddit of course: https://www.reddit.com/r/nonononoyes/comments/3qp2h7/tesla_autopilot/

All this talk about self driving cars causing accidents, very little about how they avoid them and save lives.  So here is the first video clip of a real world situation where a Tesla avoided an accident by stopping way faster than I could have.  Also note, its wet.  So if a human was driving, there would be a possibly that the human would swerve the car as well, potentially causing another accident. (click the thumbnail to start the gifv.)

ALSO, the car was smart enough not to follow the idiot driver and ‘teach them a lesson’.  Which does bring me to one concern.  How long will it be until someone decide to hack their car to do just that?  Say someone cuts you off; the car can use object recognition to track and follow the other car and then bump them off the road.

Right now, our autonomous cars do not have a voice, but I already know what to expect having grown up with Eddie Murphy comedy.

Eddie Murphy talking car. ( warning NSFW)!!!

Image from http://inhabitat.com/

Now there are exceptions to this of course, but when I tell people that in five years they will no longer need a new car, I get a look like I’m nuts.  But once I explain that you are getting a new(ish) car and its a ‘need’ not want, then they are a little less skeptical.  However, once I give the full explanation, as I’m about to here, gentle readers, you will agree with me and start telling your friends the same thing just to get that ‘your crazy’ look from them.

First, the caveats.  There are people who buy vehicles for work or hobby.  Truck, Van, taxicab or RV for example.  Then there are those that WANT to buy a new car because its a part of their identity, or love high performance, or have that classic that they love to drive.  Then there are those who live near the edge or off the grid with poor or dirt roads.

OK, now that we have the exceptions out of the way, lets talk about your typical citizen living in a urban environment. In five years, that will be the last time you need to buy a new or newish car. How does that figure?  3 things are going to start happening in 5 years and be complete in 15.

In 5 years we will have our first autonomous vehicles on the road. FULLY autonomous meaning we have the legal issues sorted out that these little Google panda cars driving around my neighborhood will no longer need a driver.  This will be huge. The idea that I can use my phone to have a car pick me up and take me where I need to go will be just as convenient as taking my own car. To make this work, there needs to be a lot of these little buggers running around such that I can have a car at my door in 3 min or less.  Assuming Uber has not been sued out of existence, they will be first in line to try and make that happen.  Watch for them to partner with google and some car maker.

In 5 years the price of these little bugs will be 1/10 the cost of a normal car.  They will be fully electric with no frills.  There will be competitors (Apple, Tesla and later other car makers) that will focus on higher end self driving cars, but those will be seen more as luxury than a necessity. The simple ‘autono-cabs’ will be simple, and thus inexpensive such that there will be many of them. So many that in 15 years many of the next generation will question why they need to ever buy a car or even get a drivers license. (By the way, when that happens we will need a new form of ID.)

Parking space, who needs a parking space?  If you live in San Francisco you know parking is a pain. Worse when you are coming home tired and the last thing you want to is run circles around the blocks trying to find a place to park.  That goes away.  Worried about finding parking when you go somewhere?  Don’t worry about it anymore.  The autono-cab will pick you up at your door and drop you off at your location. Suddenly, parking anywhere will be easier and cheaper because of the reduction of demand.  Going to the game? Or big event?  Not only will you not worry about parking, there will be special lanes for electric autono-cabs so you will get there faster then driving?  Why will this happen?  People who run events and the cities authorities where the event takes place does not want traffic anymore then you do.  They want people in and out quickly, without accidents and drunk drivers making a mess of things.  Win.

Range?  Got that covered. I looked into buying a leaf and decided not to.  If I wanted to take a day trip from mountain view to say Napa, or halfmoon bay, or Oakland (hey its kick’en over there) or even SF I could not be sure I would get back unless I was positive I could find a charging station and have the car there long enough to get a charge.  Our little cabs?  Well, being smaller, lighter and battery tech about to double the range of current car battery packs, we only need to go one way.  Odds are these little guys will have a 60-75 mile range, more then enough for a one way trip anywhere in the bay area.

Now, back to the 5 years thing.  So its 2015 (at least at the time of this writing). I just bought a 2013 PIP Prius although once the new Tesla comes out I will sell the Prius and get that.  But lets say I keep it for 5 years and buy a 2018 Tesla. (I almost never buy new). That takes us to 2020.  Many people believe, as I do, that we should start seeing autonomous cars on the road in 2020. I live in Mountain View, which is not ‘Big city’ but dense enough that the first wave of autonomous cars will reach me.  Odds are, like bike sharing, there will be a monthly or yearly fee you can pay for near unlimited rides.  So that means my Tesla stays in the garage unless I need to go somewhere outside of my local area or I want to go for a fun drive. So now that 2018 is taking on less miles and holding onto more value.  So instead of getting a new car in 2025 I can wait until 2030 or longer depending on how often I drive it.

So now we are at 2025. Battery tech will have doubled again (conservatively) making autono-cabs to more specialized public cars cheaper to make.  The autono-cab is mostly a giant tablet now with multiple screens.  Averts help reduce the costs even more. Choice increases as well as we see Apple cars, luxury Lexus and turbo Tesla’s available via an app. My 2018 Tesla is sitting in the garage most of the time and my wife got a new’ish’ car in 2015, so we have two cars that we really do not drive.  It does not make any sense to have 2 cars now, so we will sell one of them.  Now we just need to stretch the remaining car to 2030, which will be easy because again, we do not drive it much.

From www.fastcodesign.com

By 2030 the autonomous car is ubiquitous. There will still be manual cars out there but even they will have ‘safety’ overrides. The number of car accidents plummets, car insurance goes up for some, and no longer exists for others. There are gas stations, but they are also charging station for the autono-cabs.  Owning a car repair shops goes they way of the local newspaper. Traffic actually goes down instead of up even as population increases. There are more people on the road, but traffic is more efficient, almost no accidents to cause backups, no rubber necking and very few breakdowns since the cars know when they need attention and can drive themselves for service. Places like LA see a marked reduction in pollution.  And one of the biggest win IMHO, drop in drunk driving. Buying a car in 2030 does not make much sense.

Now there will be losers in this grand future I’m painting.  Vehicle drivers; cab, truck, bus will be losing their jobs. Mechanics, we will not need as many.  Oil?  Do not need as much of that either. Car insurance, car dealerships, towing services; not a growth industry.  Municipalities will need new revenue (most likely they will tax the cabs per mile). Traffic tickets?  Kiss those dollars goodbye. Toll takers?  That is already disappearing.

So lets recap.  I said that in 5 years, that will be the last time you ‘need’ to buy a new car.  Math review time!  Buy car latest in 2020.  Good for 5+ years.  We should see the first deployments of self driving public cars in 2020.  By 2022, they should cover most urban and sub-urban areas. Since you are driving your car less, you can hang on to it for another 8 years easy which brings us to 2030.  Not so crazy sounding anymore right?  1) So start telling your friends.  2) Make money bets before you explain.  3) Profit!

P.S.  I started telling people this back in May, and started writing this post in June.  So no, I’m not following the crowd, just talking WAY to long to write my posts now-a-days.

Its going to happen.  Trust me, its going to happen.

Image from http://inhabitat.com/

Ok, so this article on driver-less cabs just popped up saying NY’s Mayor de Blasio just signed with Google to have them put out a fleet of 5000 driver-less cars in Brooklyn and Queens.  They will be called Zippie.  Nice.  And that is not all; order now and get an ATM machine, vending machine spitting out mouthwash to condoms and btw the condoms will be a bit hit because I can see a trend starting on having sex in driverless cars, but that is another post.  They are even going to serve hot pizza!  What about the mess?  A 12 horse power vacuum will take care of that.  A vacumm?  How is that going to clean smeared pizza?

Does this all sound too good to be true by next year?  Yea it does.  Ok, so first off, even Google admits the driver-less car is not ready for sever weather like heavy rains, snow storms and icy roads.  This is NY we are talking about kiddies, if you ever been there in Feb.  Ug.  And serve food?  Its going to be a mess, and I’m sure some jackass will make it a worse mess.  What about legal issues?  Nothing talked about that in the post I read.

So here is what I think.. I think they are going to start putting in the ground work and doing limited tests in 2016.  Sure, you may be able to grab one of these Zippie’s next year, but I have a feeling there will be no food and running during limited times of the year ( good weather only) and limited routes.

Now you may be saying, “why such a downer man”. Actually, I’m pretty excited and looking forward to seeing even a few cabs running in limited conditions in 2016.  I want MORE cities to get in on this. Google will run the program so no one owns it. That is sooner then I would have expected from a legal perspective.  I think its great and cant wait for it.  However, lets not get crazy…. Hot Pizza?  Lets hold off on that for a bit ok?

BTW, Cabbie backlash in 3… 2… 1…