I-Like-To-Bike

Does the Tesla Full Self Driving/FSD " package" provide full self driving or not?

Or is it "just not 100% self-driving" just like my bicycle or my two older cars?

tyrion

It does.

RubeRad

Which part? cruise control is more autonomous than foot on the gas? or driver assist is more autonomous than cruise control?

I-Like-To-Bike

Your idea of a driver-less self driving vehicle is different than mine and probably any other person who isn't grasping at semantic straws to prove that driver-less cars are either here today or are ready for prime time any day now.

tyrion

No it's not.

Seattle Forrest

The problem is you guys are using facts, which I-Like is immune to.

I-Like-To-Bike

See immeditely preceding response. These features no more make a vehicle driver-less and self driving, than the convenience of the Shimano Front Freewheel Crank (FFS) feature on my old Schwinn World Tourist make that bicycle an autonomous, rider-free bicycle.

The driver/rider is still required behind the steering mechanism to get anywhere.

tyrion

That doesn't mean it's not self-driving. Currently there must be a person sitting in the driver's seat and touching the steering wheel and be ready to take over, but the car is still driving itself.

I-Like-To-Bike

OK, sure, whatever you say. I forgot, this is Foo, and words mean whatever you would like them to mean, and "facts" are whatever you want them to be.
Driver-less means what in your world?

RubeRad

I don't get what FFS does -- does it automatically shift the front for you? If so, then it may not be (fully) autonomous, but it is MORE autonomous than a manual-shift bike. Similarly, an Autobike is more autonomous than a manual-shift bike.

You are using very strict, binary definitions of 'autonomous/self-driving' for which 'more autonomous' or 'part-time self-driving' are self-negating, thus you are (mis)using terminology to beg the question, and cock-blocking any possibility of meaningful conversation.

If cruise-control is not 'more autonomous' and adaptive cruise-control (also automatically slowing down based on sensed distance to car in front) 'more autonomous' than that, and tesla autopilot (also operating the steering wheel to stay in the lane) 'more autonomous' than that, then you're the one buried in semantics.

I-Like-To-Bike

Read the subject title of this thread! DRIVER-LESS!

I-Like-To-Bike

Which version of "facts" still requires a driver in a driverless car?

RubeRad

OK fine, I didn't start this thread and write that title line. But when others of us subtly shift/expand the topic to self-driving or autonomous (by using terms such as 'self-driving' or 'autonomous') it would be cool if you could follow along.

tyrion

It means there isn't a driver. Not that complicated.

genec

The more data collected, the more test cases run, the "smarter" the system gets... that does not happen with human drivers... each human starts at square 1... and has to learn everything all other drivers have learned before them.

Intelligence in AI driving improves daily... how many "shopping cart ladies" will killed in the future... or how many road crossing tractortrailers will be hit in the future by autonomous vehicles? Meanwhile, humans continue to "right cross" cyclists... year after year. AI gets smarter, humans do not. (Ultimately that may be the downfall of humans... but that's a different thread.)

Sensors improve... human sight and reaction times do not. In fact as each human ages, we get worse... after going through a period of steady improvement... us frail humans decline in performance.

AI is likely to be shared, if through no more than a greedy system of IP licenses.

fmoir

ford is going all electric by 2030

fmoir

True that!

tyrion

Not only does the autopilot recording dataset constantly grow, so does the learning technology that learns from that dataset. Another double whammy leverage that humans don't have.

Seattle Forrest

While you're playing word games, everyone around you is having an intelligent conversation.

I-Like-To-Bike

Note that electric motors, like cruise control, do not make a vehicle driver less.

I-Like-To-Bike

So just how smart and learned have Google, Tesla, Uber, Ford, et al. become? Perhaps smart enough to recognize that they are a long way off from being able to move beyond testing their proprietary prototypes in a few limited carefully selected geo-fenced zones. The more recent gossip from the PR flaks indicates that the introduction date for production and sale of driverless cars keeps getting pushed out into the future for fielding real driverless cars (ya know, SAE Level 5 autonomous cars, not cars with advanced cruise control.)

It seems only in Internet gossip threads like this and Elon Musk Twitter-dom is there anyone seriously suggesting that real driverless cars will be available for use by the public in the next 4 or 5 years. And that speculation is fueled more by wishful thinking, references to data collection metrics, and redefining driverless cars as something that still requires constant monitoring by a human driver, than real honest-to-goodness facts

Good take on the subject is at: The rhetoric about driverless cars is being toned down

genec

And from the reader responses...
https://wreg.com/2020/01/02/multi-ca...lane-closures/
https://www.reviewjournal.com/local/...uries-1926386/
https://www.cnn.com/2019/11/26/weath...rnd/index.html


Silly human drivers... they just *think* they can drive in any and all conditions...

I-Like-To-Bike

Yep, sure, Google knows better from their testing regimen and data collection that there is no reason to field a car that the passenger might need to use to venture beyond a few miles surrounding Chandler AZ or in any weather or road or traffic condition that varies too much from what can be found there. Not that Google has any intention to manufacture or field any cars, anywhere. Data collection, map making, and selling collected data, software and ads is the name of their game.

Uber would like to have self driving DRIVERLESS cars for the obvious monetary reason of pocketing the entire fare and not paying those pesky drivers a dime. Perhaps they expect to take ownership of their future DRIVERLESS taxis also without giving a dime to whomever will build them. Maybe Uber might even make a dime of profit if venture capitalists like Soft Bank will give Uber/Lyft Billions of Dollars to buy the fleet of (vaporware) driverless taxis with little to no expectation of ever getting repaid (just like the Ponzi-WeWorks fiasco.)

However, I seriously doubt that all the Billions of Dollars being burned on these prototype testing endeavors by Google, Uber, Ford, Volvo, M-B, et al. are for the purpose of producing an end product with fancy cruise control, built-in Internet connection, and proximity warning signals, BUT still requires a driver behind the wheel of every one of the vehicles when in motion. Won't be many cheers from the speculators and venture capitalists if that is all they get for their massive expenditures on this project. Sorta like settling for Tang instead of a moon landing and safe return.

My prediction is that if Alphabet/Google cannot monetize their mapping and obstacle detecting software being developed by Waymo and license or sell it for significant money in the next 4 years to some other organization that believes they can profitably utilize it in production vehicles, Alphabet/Google/Waymo will wash their hands of the entire Waymo money burning autonomous car project.

clemsongirl 

i find it funny how this thread deteriorated into such unreasonableness and chest beating. Someday level 5 autonomous cars will arrive in large numbers and the serious crashes that are due to human error and distracted, speeding and drunken driving will be much less and later none. For now i’ll just enjoy the safety features that the research/development into it have already and will continue to give us.

genec

Or we simply let the Chinese beat us at this technology too... https://www.businesswire.com/news/ho...-M1-Officially

Why? Because they WANT to...