Tesla's Autopilot

[Mediocre]

This is how a Tesla automobile sees the road. And with the upcoming advanced hardware in 6 months and the rollout of 5g technology in 2020, the transition is almost complete.

https://www.youtube.com/watch?v=7ztK5AhShqU

For the video
https://electrek.co/2018/10/15/tesla-new-autopilot-neural-net-v9/amp/

https://kenyatalk.s3.amazonaws.com/2018/10/265082_072898021b94b738b2de65d424b827c4.jpg
For the image
https://electrek.co/2018/10/16/tesla-neural-net-computer-production-elon-musk/amp/

[kah_tony]

You can never replace human intuition with AI unless you end up creating a human.

Scenario: A vehicle is approaching an intersection where a young boy is running across, there are three alternatives:

1. Go to opposite lane and hit on coming traffic.
2. RUN OVER the child
3. Hit the barrier or go over the ledge and only driver is injured or killed

Kazi kwako Elon Mutiso.

[Mediocre]

@kah tony

You can never replace human intuition with AI unless you end up creating a human.

Scenario: A vehicle is approaching an intersection where a young boy is running across, there are three alternatives:

1. Go to opposite lane and hit on coming traffic.
2. RUN OVER the child
3. Hit the barrier or go over the ledge and only driver is injured or killed

Kazi kwako Elon Mutiso.

Umecheki video? The algorithms can measure the speed of any moving object relative to the car they are driving. With 8 cameras in all directions taking in 200 - 2000 frames per second at HD resolution they can see the child coming way before you know it. They’ll just slam the breaks on regardless of whether ulikuwa umeweka gear 5. It’s called a hardware interrupt.

[kah_tony]

Iv

@Mediocre

Umecheki video? The algorithms can measure the speed of any moving object relative to the car they are driving. With 8 cameras in all directions taking in 200 - 2000 frames per second at HD resolution they can see the child coming way before you know it. They’ll just slam the breaks on regardless of whether ulikuwa umeweka gear 5. It’s called a hardware interrupt.

I have seen the video. Jua that an algorithm is only efficient as its programmer. Possibilities are infinite while the computer is definitely finite.

[Mediocre]

@kah tony

Iv

I have seen the video. Jua that an algorithm is only efficient as its programmer. Possibilities are infinite while the computer is definitely finite.

This is not a routine based programming, this is an algorithm that learns about it’s surroundings better and faster than a human being. That’s why inaitwa artificial intelligence.

[shocks]

@kah tony

You can never replace human intuition with AI unless you end up creating a human.

Scenario: A vehicle is approaching an intersection where a young boy is running across, there are three alternatives:

1. Go to opposite lane and hit on coming traffic.
2. RUN OVER the child
3. Hit the barrier or go over the ledge and only driver is injured or killed

Kazi kwako Elon Mutiso.

There is a podcast I think by radio lab that looked at this. Since it relies on algorithms, it has to be told what to do. And who should make such a decision, the motor company, the insuarance, law makers in each country? In 2016, team leader wa the self driving mercedes project said their car is programmed to always protect the occupants, you guessed it, it raised a shit storm.
Swali ngumu, yet nyeti

[Wechez]

Elon Musk is a fraud.

[Mediocre]

@shocks

There is a podcast I think by radio lab that looked at this. Since it relies on algorithms, it has to be told what to do. And who should make such a decision, the motor company, the insuarance, law makers in each country? In 2016, team leader wa the self driving mercedes project said the car is programmed to always protect the occupants, you guessed it, it raised a shit storm.
Swali ngumu, yet nyeti

That’s a loaded statement. The algorithms are designed to minimize error and prevent any contact with another object or entity while driving. You can’t program for ‘protect occupant’ since its not scientifically quantifiable.

[Rene_Descartes]

Machines can do these things, but unless humans become machines so that they can fit in a world of machines they will always be a novelty.why do cars such as the range rover which is more than 5 tons still have a future carrying a load of 200kg? because we are humans.

[emali]

@shocks

There is a podcast I think by radio lab that looked at this. Since it relies on algorithms, it has to be told what to do. And who should make such a decision, the motor company, the insuarance, law makers in each country? In 2016, team leader wa the self driving mercedes project said their car is programmed to always protect the occupants, you guessed it, it raised a shit storm.
Swali ngumu, yet nyeti

The so called “trolley problem”

[shocks]

@emali

The so called “trolley problem”

It’s no longer theoretical

[123tokambio]

@Wechez

Elon Musk is a fraud.

In what way sir, if i may ask…

[123tokambio]

@kah tony

You can never replace human intuition with AI unless you end up creating a human.

Scenario: A vehicle is approaching an intersection where a young boy is running across, there are three alternatives:

1. Go to opposite lane and hit on coming traffic.
2. RUN OVER the child
3. Hit the barrier or go over the ledge and only driver is injured or killed

Kazi kwako Elon Mutiso.

You have three, it probably has 7, including deploying an outside airbag to cushion the pedestrian in case of a collusion…yes, its coming…

[Bark]

@kah tony

You can never replace human intuition with AI unless you end up creating a human.

Scenario: A vehicle is approaching an intersection where a young boy is running across, there are three alternatives:

1. Go to opposite lane and hit on coming traffic.
2. RUN OVER the child
3. Hit the barrier or go over the ledge and only driver is injured or killed

Kazi kwako Elon Mutiso.

That’s an ethical dilemma not even humans can come up with a “right” answer for. However, a subconscious alertness that technology provides sounds better than human intuition in most situations.