Are you on Team Bollinger or Team Rivian?

[JSH]

Originally Posted by cRiPpLe_rOoStEr

Sure it's a dishonest approach, but since Tesla is taking benefits from such strategy it might explain why it didn't come to an end (yet).

True. FSD has an infinite profit margin since they are charging buyers for something that requires no additional hardware and doesn't exist.

[rmay635703]

Originally Posted by cRiPpLe_rOoStEr

Even for commercial applications, I'd still not expect most people to approve it. Either for some skepticism about its efficiency or simply because they would rather indulge themselves pretending to be somewhat "superior" to a driver.

As recent events re-enforce it doesn’t matter what people want
They will take it and they will like it ready or not.

[RedDevil]

Do we learn from the past or not?
We'll never have infinite access to information that fits in our pockets.
We'll never fly to the moon.
Complex, noisy and expensive cars will never replace horses.
Flimsy houses are no comparison to a solid cave.

The learning behind FSD is progressing to another level. While the data set of traffic situations is growing rapidly, so is the success rate of the decisions the neural network makes. Once the latter increases faster than the former the quality of the decisions will accelerate and surpass human abilities.

For comparison, as a software engineer and amateur Go player I have been working on a Go playing program in the 90s. Our program never exceeded the beginner level while playing the small 9x9 board, but we had hope of challenging the strongest programs of the time which I could give a 5-7 stone handicap advantage and still beat, not being that strong myself.
Over time the programs slowly got better, approaching and finally surpassing my level in the early naughties, but they were still no match for the stronger amateur players, never mind professional Go players.
It was generally accepted that Go was too big a challenge for AI.

Then the AlphaGo team got rolling and quickly surpassed even the strongest Go players, each iteration completely destroying the previous one in playing strength.
The current version, AlphaZero, has a crazy high ELO rating. It is absolutely obvious that no human player will ever come near that level.

Once the learning gets more refined the power needed to get there diminishes, so instead of using a purpose-built supercomputer you can now have top professional strength from a single PC with Leela Zero.

I expect the same will happen with FSD. The neural net will keep refining and condensing to a limited set of really important rules, reducing the computational power needed for each level, allowing for raising the bar continuously.

[cRiPpLe_rOoStEr]

Originally Posted by RedDevil

I expect the same will happen with FSD. The neural net will keep refining and condensing to a limited set of really important rules, reducing the computational power needed for each level, allowing for raising the bar continuously.

The amount of variables might lead FSD to require some level of precision which hasn't been applied not even to military equipments or commercial aircraft. Even if it succeeds someday, it's too far from being able to deal with all the driving conditions which might be experienced by humans. Under certain circumstances, I'd still trust someone driving under influence of alcohol before I try a robot-car.

[freebeard]

tech.slashdot.org: Tesla Model 3 Drives Straight Into Overturned Truck In What Seems To Be Autopilot Failure (jalopnik.com)

Originally Posted by TFA

Tesla's Autopilot system primarily uses cameras for its Autopilot system, and previous wrecks have suggested that situations like this, a light-colored large immobile object on the road on a bright day can be hard for the system to distinguish. In general, immobile objects are challenging for emergency automatic braking systems and autonomous systems, as if you use radar emitters to trigger braking for immobile objects, cars tend to have far too many false positives and unintended stops than is safe or desirable.

[JSH]

Originally Posted by freebeard

tech.slashdot.org: Tesla Model 3 Drives Straight Into Overturned Truck In What Seems To Be Autopilot Failure (jalopnik.com)

That is one of the reasons LIDAR is critical for self-driving.

I found this part of the article shocking:

"driving Tesla was a 53-year-old man named Huang, who claimed to have turned on the vehicle assist system at the time. It was thought that the vehicle would detect an obstacle and slow down or stop, but the car still moved at a fixed speed"

The guy claims he saw the truck but just waited for the car to stop on it's own instead of manually applying the brakes.

[Piotrsko]

Sounds like a default system for dealing with odd things needs to be installed so that the car just stops / moves over, whatever. Currently, it just kinda goes catatonic like a buffer overflow on your PC

[Hersbird]

Seems to me for self driving to work...
Step 1: stop car before running into things.
Let me know when they get step one down and we can talk about the next step.

[RedDevil]

10 billion kilometers, you'll find a couple of bizarre things.

Don't people drive into things? Yes they do, 5 times as often as FSD does.

FSD gets better every day. People do not.

[freebeard]

Originally Posted by Piotrsko

Sounds like a default system for dealing with odd things needs to be installed so that the car just stops / moves over, whatever.

The Book of Proverbs and Ray Stevens agree: "None so blind as he who will not see."