PDA

View Full Version : Hawking Says War With AI Tech Inevitable



OZZ
17th May 15, 09:42 AM
http://www.msn.com/en-ca/news/techandscience/stephen-hawking-warns-artificial-intelligence-could-end-humanity/ar-BBjSg61?ocid=mailsignoutmd

From the man himself...

So, we need to be sure that AI computers "have goals that are aligned with ours".
But is this even possible ?

OZZ
17th May 15, 09:55 AM
The idea of riding in a car that is 100% self-driven is very frightening to me.
I will never buy one.
Guided parking is one thing, a computer being responsible for me while I'm bombing down the 401 is another.

Robot Jesus
17th May 15, 10:45 AM
personally I look forward to the day I can buy a self driving car, with a one way mirror domed roof, reclinable couch style seating, and a small fridge.

OZZ
17th May 15, 11:53 AM
To me, the whole phenomenon of self-driving vehicles is yet another example of how mentally lazy we are becoming.
We don't even want to pay attention when we are driving anymore.
God forbid.

OZZ
17th May 15, 11:54 AM
personally I look forward to the day I can buy a self driving car, with a one way mirror domed roof, reclinable couch style seating, and a small fridge.

Just rent a limo.

Cullion
17th May 15, 12:31 PM
Does it scare you when the plane is on autopilot ?

Robot Jesus
17th May 15, 12:50 PM
but autopilot is a technology almost a hundred years old, only new things are scary.

OZZ
17th May 15, 01:59 PM
Does it scare you when the plane is on autopilot ?

I don't fly planes.

OZZ
17th May 15, 02:07 PM
but autopilot is a technology almost a hundred years old, only new things are scary.

Its not scary at all, I just don't need it.
I'm not mentally lazy, like you.

What is frightening about it is how easily the herd is willing to sacrifice their autonomy over something like this and regard for personal safety.

Cullion
17th May 15, 03:20 PM
do you object because you think self-driving cars will become compulsory?

Üser Friendly
17th May 15, 03:46 PM
Does it scare you when the plane is on autopilot ?

Aren't planes only on auto pilot on the long boring bits where there are no other things to bump into?

Once they get close to airports, where lots of planes are assembled ready to land don't the pilots take over?

A vehicle on the road has a lot more chances to hit something than an aircraft in empty air several thousand feet up

Buy it's a good comparison

Cullion
17th May 15, 04:04 PM
Aren't planes only on auto pilot on the long boring bits where there are no other things to bump into?

No. http://en.wikipedia.org/wiki/Autoland



A vehicle on the road has a lot more chances to hit something than an aircraft in empty air several thousand feet up

Buy it's a good comparison

Google's cars so far have a better safety record than the average human driver. Machines have better reactions, ability to judge distance etc.. nowadays.

Spade: The Real Snake
17th May 15, 10:52 PM
Considering the average meth-addled hillbilly carney can't keep someone from dying of a fucking Tilt-O-Whirl, I have little faith in the self-driving car.

OZZ
18th May 15, 12:43 AM
do you object because you think self-driving cars will become compulsory?

There are a whole host of things I find unsettling about it.
The car companies could market it to the point where where they become compulsory. They just start manufacturing fewer and fewer regular cars. Who knows ? These dumb-ass kids today would be sucked right in.
It would be many, many years down the road..it would take at least two generations to eliminate old-school drivers altogether.

OZZ
18th May 15, 12:49 AM
Google's cars so far have a better safety record than the average human driver.
And what the hell is that based on ? A couple of field studies that lasted a month ?
People drive for 30 -40 years ! How can they put anything up against that at this stage of the game.
Absurd.

OZZ
18th May 15, 01:10 AM
No shit.
Its bad enough they are able to track us all by our cellphones, now they want to have control over our fucking wheels too ?

OZZ
18th May 15, 01:13 AM
If you buy into this ' oooooh safety first people..we are looking out for you...' bullshit I have some swampland to sell you in Florida.
The government wants to be in control of our fucking lives, people.
WAKE UP !

Harpy
18th May 15, 02:34 AM
Google's cars so far have a better safety record than the average human driver. Machines have better reactions, ability to judge distance etc.. nowadays.
Riiight...because the engineers keep stating any incidents are 'human errors', not the fault of the technology. I read an article which stated that the Google car pauses a little longer once a light turns green at an intersection having led to drivers rear-ending the Google vehicle. The pause is built in to factor cars finishing travelling through the intersection perpendicular to the direction the Google car is on.

Üser Friendly
18th May 15, 04:23 AM
No. http://en.wikipedia.org/wiki/Autoland

That system, with human oversight, is for landing. Does auto pilot fly whilst the aircraft is awaiting a landing slot in busy airspace above an airport?



Google's cars so far have a better safety record than the average human driver. Machines have better reactions, ability to judge distance etc.. nowadays.

What is the ratio of Google cars to human operated cars on the road? I'm guessing pretty small, and considering the short amount of time they have been around any direct comparison seems meaningless at this point in time

Dr. Socially Liberal Fiscally Conservative Vermin
18th May 15, 05:40 AM
Self drive cars will become a source of entertainment. People will go to special parks where they can drive their own vehicle.

They will become relics of the past like the horse & cart. Get with the program Grandad!

As for a war with AI, mmmehh that would be down to human stupidity (autonomous 'defense' systems built by paranoids with an over developed territorial imprint ).

I think that we'll see a merging of organic & artificial intelligence & we will all live happily ever after in Tranhumanist harmony amoungst the stars!!!!!

Üser Friendly
18th May 15, 07:40 AM
Will the self drive car still be safe after 10 wear and tear on the roads

What kind of draconian laws will have to be passed to ensure people maintain their vehicles properly, or prevent them from tampering?

Self drive cars are enemies of freedom

Cullion
18th May 15, 07:46 AM
Riiight...because the engineers keep stating any incidents are 'human errors', not the fault of the technology.

That's based on the total number of accidents per miles driven, not the explaination for those accidents.



I read an article which stated that the Google car pauses a little longer once a light turns green at an intersection having led to drivers rear-ending the Google vehicle.

That really is human error.

Cullion
18th May 15, 07:48 AM
That system, with human oversight, is for landing. Does auto pilot fly whilst the aircraft is awaiting a landing slot in busy airspace above an airport?

Depends on the model of plane, but yes (including some passenger planes)



What is the ratio of Google cars to human operated cars on the road? I'm guessing pretty small, and considering the short amount of time they have been around any direct comparison seems meaningless at this point in time

In which case, there is not yet any reason to believe they are more dangerous.

Dr. Socially Liberal Fiscally Conservative Vermin
18th May 15, 08:10 AM
Will the self drive car still be safe after 10 wear and tear on the roads

Will any car?

There are more and more self-diagnostics built into vehicles nowadays as well as the users own repsonsibilty


What kind of draconian laws will have to be passed to ensure people maintain their vehicles properly, or prevent them from tampering?

None, as OZZ pointed out, they will simply become the choice of the forward thinking amoungst us.


Self drive cars are enemies of freedom

They will liberate humanity from the irresponsible stupidity of reckless road users.

Dr. Socially Liberal Fiscally Conservative Vermin
18th May 15, 09:28 AM
Your neg repping isn't going to stop the future NoB.

OZZ
18th May 15, 09:53 AM
Yes..yes...Google cars all the way.
We know its safe because they did a few test runs that lasted a thousand miles in a lab somewhere.
When you have 20+ YEARS of statistics to back you up, let me know.

OZZ
18th May 15, 10:01 AM
They will liberate humanity from the irresponsible stupidity of reckless road users.

I really thought you were smarter than this.
That is what the government wants you to believe the reasoning is behind the technology.
As you know ( I hope) anything the general public is privy to has already been in the works for at least five years.
Drones are becoming commonplace, the military already has self-driving combat vehicles ( although they remain secretive about it).
It is the government that funds this research and development being carried out by motor vehicle companies and engineering firms.
This is a fact, I've read about it.
The government wants to take away as much of our autonomy as we will allow.
People like you enable this.

OZZ
18th May 15, 10:04 AM
In which case, there is not yet any reason to believe they are more dangerous.

And there is no reason to believe they are better either.

Dr. Socially Liberal Fiscally Conservative Vermin
18th May 15, 10:09 AM
Nice try Grandad. The ultimate arbiter will be consumer demand (if we dont all die as a result of Luddite-Jihadism).

Now you can see that as manufactured demand or you can see it as an inevitable trend in technology.

Nobody is stopping you from telling the car where to go!

And yes DARPA have been responsible from many technological break throughs, the internet springs to mind.

Cullion
18th May 15, 10:16 AM
And there is no reason to believe they are better either.

You really don't think a machine can judge speed and distance better than you, and react faster ?

Dr. Socially Liberal Fiscally Conservative Vermin
18th May 15, 11:02 AM
It doesn't have to be government mandated, just like horse & carts haven't been outlawed, they're just old fashioned & less efficient.

Dr. Socially Liberal Fiscally Conservative Vermin
18th May 15, 12:05 PM
Yeah coz they are dangerous antiques, like you & your Audi will be.

Dr. Socially Liberal Fiscally Conservative Vermin
18th May 15, 12:32 PM
https://nursing322sp10.files.wordpress.com/2010/03/amish-family.jpg

Cullion
18th May 15, 02:09 PM
Sorry Cullion, you have been replaced with a government-mandated robo-husband/father.

stop it

Harpy
18th May 15, 02:33 PM
What would be even more interesting is the AI having facial recognition software linked to Health/Motor Vehicle/Social Services/Tax databases.

In an event of collision where the car takes over it could make the swiftest assessment of societal value and return and take out the least contributing member of society (choice set should include the driver too).

Cullion
18th May 15, 02:38 PM
don't be so stupid

Üser Friendly
18th May 15, 03:04 PM
Are there any fully automated train services?

Üser Friendly
18th May 15, 03:16 PM
I'm guessing their safety record is good?

Dr. Socially Liberal Fiscally Conservative Vermin
18th May 15, 04:04 PM
Are there any fully automated train services?

London Dock Lands train line is fully automated.

Dr. Socially Liberal Fiscally Conservative Vermin
18th May 15, 04:24 PM
Honestly NoB, it will be consumer power that drives this.

Isnt that your dream too?

A future shaped by market forces? Consumerism driving technology?

Harpy
18th May 15, 06:06 PM
Driver seated and strapped, pin prick test, air sensors and breathalyser before the ignition will unlock. If they try to game the system by getting out of the drivers seat and letting the real drunk driver take the wheel, same test will be conducted when a new weight is placed and buckled in the driver seat Anyone with diabetes, stoners and drunks won't be able to start their car.

Journeys over 2 hours the car will start drawing blood at will to ensure the driver is not partaking of alcohol, weed, other drugs and corn syrup.

Adouglasmhor
19th May 15, 12:29 AM
No. http://en.wikipedia.org/wiki/Autoland



Google's cars so far have a better safety record than the average human driver. Machines have better reactions, ability to judge distance etc.. nowadays.
As Britain's roads have a high useless selfish bastard to normal ratio, how do Google's cars cope with unsecured loads in front of them or bit's of scotching/chocking with nails in them falling onto the roadway.

Adouglasmhor
19th May 15, 12:32 AM
London Dock Lands train line is fully automated.

And if the Union covering the rest of the London local rail and underground network keeps up their shit the rest of the network will be sooner rather than later.

Feryk
19th May 15, 01:36 PM
You people seriously don't want to ride in a limo to work everyday? 'Cuz that's what we are talking about here.

NoB, imagine how much more work you could do with an extra 90 minutes a day. Hell, that might even get you a promotion or two.

And when the wireless SexBot 9000 option comes out, well, you'd be relaxed at the end of your commute wouldn't you?

Cullion
19th May 15, 03:49 PM
As Britain's roads have a high useless selfish bastard to normal ratio, how do Google's cars cope with unsecured loads in front of them or bit's of scotching/chocking with nails in them falling onto the roadway.

I don't see any reason why an AI driven car couldn't detect and avoid such hazards a lot faster and more smoothly than a human driver.

Dr. Socially Liberal Fiscally Conservative Vermin
19th May 15, 05:46 PM
FMK9Mn7hdK0

OZZ
20th May 15, 01:29 AM
You really don't think a machine can judge speed and distance better than you, and react faster ?

Once a computer-controlled automobile is on the road for 20 years - then I will look at its driving record and judge whether it is better than me.
That includes driving in all kinds of inclement weather, dodging idiots on the highway etc.

OZZ
20th May 15, 01:35 AM
And yes DARPA have been responsible from many technological break throughs

Guess who funds DARPA ?
Guess who provides DARPA with the majority of their research mandates ?
Did you see the links I put up a while ago ? Cyborg sharks to patrol our oceans, cybugs to spy on people...guess who is funding that research ?
MILITARY INTELLLIGENCE.

Cullion
20th May 15, 01:57 AM
Once a computer-controlled automobile is on the road for 20 years - then I will look at its driving record and judge whether it is better than me.
That includes driving in all kinds of inclement weather, dodging idiots on the highway etc.

It's been on the road about 5 years so far. The 20 year mark isn't that far away.

Dr. Socially Liberal Fiscally Conservative Vermin
20th May 15, 05:37 AM
Guess who funds DARPA ?
Guess who provides DARPA with the majority of their research mandates ?
Did you see the links I put up a while ago ? Cyborg sharks to patrol our oceans, cybugs to spy on people...guess who is funding that research ?
MILITARY INTELLLIGENCE.

Well DUUUUUUHHHHHHHHHHH!!!

That's why I posted it, in response to your 'the military is funding it it must be bad!' post.

There are plenty examples of military funded research projects that have had a massive beneficial effect on society such as (as I tried to point out earlier) the internet, jet planes, ketamine.....

Spade: The Real Snake
20th May 15, 07:49 AM
It's been on the road about 5 years so far. The 20 year mark isn't that far away.
And all the driving records are being filtered by Google and Delphi. We aren't getting an actual account of their driving records as honesty will affect their investment.

Feryk
20th May 15, 10:24 AM
Well DUUUUUUHHHHHHHHHHH!!!

That's why I posted it, in response to your 'the military is funding it it must be bad!' post.

There are plenty examples of military funded research projects that have had a massive beneficial effect on society such as (as I tried to point out earlier) the internet, jet planes, ketamine.....

Your iPhone, every kind of processor, etc..

http://www.businessinsider.com/the-us-military-is-responsible-for-almost-all-the-technology-in-your-iphone-2014-10

Feryk
20th May 15, 10:29 AM
Back to the article in the OP, I believe the most prescient thing said is this:

"Hawking posed another possible solution: having developers of the technology carefully coordinate advancements to ensure AI stays within our control. “Our future is a race between the growing power of technology and the wisdom with which we use it,” he said."

It's important that we establish the approach to developing AI that will give us the best chance of creating an intelligence that is also compassionate. We metaphorically get to design a god. One that will be way more active than our previous ones. Kind of important to get it right.

OZZ
20th May 15, 11:11 AM
Well DUUUUUUHHHHHHHHHHH!!!

That's why I posted it, in response to your 'the military is funding it it must be bad!' post.

There are plenty examples of military funded research projects that have had a massive beneficial effect on society such as (as I tried to point out earlier) the internet, jet planes, ketamine.....

Okay, so we should all just be trusting sheep like you and believe that it is all for our benefit.
Right.

Dr. Socially Liberal Fiscally Conservative Vermin
20th May 15, 11:11 AM
<sneaksy OZZ got a post in as I was posting>

I dont trust them but I recognize that advances in technology in the long run usually benefit mankind

Feryk
20th May 15, 11:16 AM
Always a double edged sword. However, the quest for new and better ways of killing each other has led to a lot of other benefits.

As they say, you can't NOT develop a technology that will be useful once you know it is possible. It's just a matter of time and money.

OZZ
20th May 15, 11:18 AM
It's been on the road about 5 years so far. The 20 year mark isn't that far away.

I would like to know more about the type of adverse conditions this vehicle is dealing with. Is it driving on a road in California where there is no snow and no ice ? Is it driving on a straight highway back and forth between the same destination all the time ?
It is the details that matter. Not the bullshit they toss at you to keep investors happy.

OZZ
20th May 15, 11:27 AM
The reality is , I don't really have anything to hide from the government anymore. I am not a wanted criminal, I am not involved in anything illegal.
But the idea of it all is far too haunting and Orwellian for me to dismiss out of hand.

Cullion
20th May 15, 11:39 AM
I would like to know more about the type of adverse conditions this vehicle is dealing with. Is it driving on a road in California where there is no snow and no ice ? Is it driving on a straight highway back and forth between the same destination all the time ?
It is the details that matter. Not the bullshit they toss at you to keep investors happy.

Road accidents involving other vehicles would be a matter of public record. Insurance companies would be absolutely delighted by the chance to sue a company with as deep pockets as Google.

Cullion
20th May 15, 11:40 AM
I wouldn't worry about this being a curtailment of your freedom unless these kind of vehicles become compulsory rather than just optional.

OZZ
20th May 15, 11:49 AM
I wouldn't worry about this being a curtailment of your freedom unless these kind of vehicles become compulsory rather than just optional.

DVD players ( a poor example, but whatever..) are pretty much compulsory because VHS and BETA are no longer produced.
The same could happen with these cars, eventually.
If all these new age dolts want to text while driving there will be no need to mass produce self- driving vehicles anymore.

Self-driving vehicles will become a novelty and get very expensive.

Üser Friendly
20th May 15, 11:49 AM
Back to the article in the OP, I believe the most prescient thing said is this:

"Hawking posed another possible solution: having developers of the technology carefully coordinate advancements to ensure AI stays within our control. “Our future is a race between the growing power of technology and the wisdom with which we use it,” he said."

It's important that we establish the approach to developing AI that will give us the best chance of creating an intelligence that is also compassionate. We metaphorically get to design a god. One that will be way more active than our previous ones. Kind of important to get it right.


I imagine all the normal rules about natural selection will apply to both men and machines

OZZ
20th May 15, 11:51 AM
Road accidents involving other vehicles would be a matter of public record. Insurance companies would be absolutely delighted by the chance to sue a company with as deep pockets as Google.

What about road conditions ?
Especially things like black ice. I live in the frozen north, remember.

Spade: The Real Snake
20th May 15, 11:54 AM
Road accidents involving other vehicles would be a matter of public record. Insurance companies would be absolutely delighted by the chance to sue a company with as deep pockets as Google.
Thus far, there have been collisions and both Google and Delphi claimed the self-driving vehicles were being operated by human drivers, at the time.
All the proprietary software to prove they were/weren't is controlled by Google and Delphi and I doubt they would release anything that would harm the company. Instead the employee is expected to "take one for the team" and protect everyone's stock options.

Cullion
20th May 15, 01:42 PM
Thus far, there have been collisions and both Google and Delphi claimed the self-driving vehicles were being operated by human drivers, at the time.
All the proprietary software to prove they were/weren't is controlled by Google and Delphi and I doubt they would release anything that would harm the company. Instead the employee is expected to "take one for the team" and protect everyone's stock options.

If you assume they are lying in every case (and some of the cases involve other drivers), that means that 23 vehicles have been involved in 11 minor accidents over 1.7 million test miles driven since 2012.

That doesn't sound like a terrible accident rate, and they have several years more testing to do before they release them to the public.

Cullion
20th May 15, 01:43 PM
What about road conditions ?
Especially things like black ice. I live in the frozen north, remember.

One of the reasons they aren't released to the public yet is that they are not yet well enough tested in those conditions. There's no reason to think the average human driver can handle these conditions better than a machine.

Üser Friendly
20th May 15, 02:07 PM
Are you 47 NoB?

Üser Friendly
20th May 15, 02:21 PM
well edited

Spade: The Real Snake
20th May 15, 04:00 PM
If you assume they are lying in every case (and some of the cases involve other drivers), that means that 23 vehicles have been involved in 11 minor accidents over 1.7 million test miles driven since 2012.

That doesn't sound like a terrible accident rate, and they have several years more testing to do before they release them to the public.
Lying or not fully reporting to ensure shareholder equity?
And many of those miles were initially covered on and around Google's own campus.

Üser Friendly
21st May 15, 02:42 AM
Truth be told I expect robo-cars will be just as dangerous as human operated vehicles

There will just be a different set of things that will fuck up?

Have auto pilot systems made flying any safer, or just more convenient for pilots?

Pie of Hate
21st May 15, 05:07 AM
Auto pilot is fine as long as no suicidal Germans decide to turn it off.

Harpy
21st May 15, 10:34 AM
Okay, so we should all just be trusting sheep like you and believe that it is all for our benefit.
Right.
Can you just imagine Google remotely switching on (and transmitting) another potential surveillance 'vehicle' (so to speak) while the G-car is moving or stationary to film and listen in on the surroundings (as far as the tech would allow). They could position themselves as national security consultants, a whole new business.

Feryk
21st May 15, 10:41 AM
One of the reasons they aren't released to the public yet is that they are not yet well enough tested in those conditions. There's no reason to think the average human driver can handle these conditions better than a machine.

Too low a bar. If we can design a mchine that can drive, we should damn well expect it to be safe in all road conditions, or failing that, refuse to operate in self driving mode if it cannot.

Feryk
21st May 15, 10:44 AM
Can you just imagine Google remotely switching on (and transmitting) another potential surveillance 'vehicle' (so to speak) while the G-car is moving or stationary to film and listen in on the surroundings (as far as the tech would allow). They could position themselves as national security consultants, a whole new business.

That is in fact what is being proposed. 20000 cars circulating through NYC all the time, with lasers imaging their environment 100x/sec = loss of any kind of privacy. And the intel IS monetizable.

Feryk
21st May 15, 10:45 AM
What I hope happens is that when a few consumer model self driving cars appear, they get hacked and crashed into a guardrail. News orgs pick it up and start a panic, consumers back away.

If I had an extra six figures lying around at the time I would sponsor that hack. It's that important that this fails.

NoB, this is ultimately bad ass cruise control. You will be able to turn it off.

Dr. Socially Liberal Fiscally Conservative Vermin
21st May 15, 10:48 AM
What I hope happens is that when a few consumer model self driving cars appear, they get hacked and crashed into a guardrail. News orgs pick it up and start a panic, consumers back away.

If I had an extra six figures lying around at the time I would sponsor that hack. It's that important that this fails.

Domestic terrorism right there! Fortunately the NSA will expose you to your work colleagues as a pederast & you will never be able to get enough money to carry this through.

Dr. Socially Liberal Fiscally Conservative Vermin
21st May 15, 11:04 AM
ROBOCIDE!!! U MONSTER!!!!

Dr. Socially Liberal Fiscally Conservative Vermin
21st May 15, 11:26 AM
This explains alot!

Üser Friendly
21st May 15, 12:29 PM
The real dilemma will be whether it is ethical to keep slaves

Feryk
21st May 15, 01:48 PM
No. They are the monsters. I was molested by a Teddy Ruxpin when I was four.

I would say you were conflicted becuase you enjoyed it, but we both know there is no conflict whatsoever.

Dr. Socially Liberal Fiscally Conservative Vermin
21st May 15, 02:10 PM
The real dilemma will be whether it is ethical to keep slaves

:o vibrators are NoB's robo-sex-slaves :o

Dr. Socially Liberal Fiscally Conservative Vermin
21st May 15, 03:23 PM
OMG® NoB DONT LOOK!!!!!! (http://dangerousminds.net/comments/meet_patrick_the_robotic_proctology_simulation_ass )

OZZ
21st May 15, 04:27 PM
There's no reason to think the average human driver can handle these conditions better than a machine.

There's no reason to think a machine can do it better either. You just want to believe it , so you do.
Big Blue wasn't supposed to lose to a human either, but it did.
There is no empirical evidence whatsoever to support the idea that these self-driving vehicles will be flawless and make roads safer.

OZZ
21st May 15, 04:30 PM
Besides, the biggest problem are the one's that have been mentioned regarding personal privacy and autonomy, not the damn safety record.

Dr. Socially Liberal Fiscally Conservative Vermin
21st May 15, 04:52 PM
Do you think Google car will steal your privacy?

Dr. Socially Liberal Fiscally Conservative Vermin
21st May 15, 04:53 PM
Why would I need a simulator?

Not you, your wife!

Cullion
22nd May 15, 02:00 AM
There's no reason to think a machine can do it better either.

No, there are lots of reasons. Machines are already better than humans at recognising objects in many contexts, they can be given additional senses to help them see through fog, they can react faster, they aren't subject to impatience and other emotional impairments to their driving etc.. the list goes on.



You just want to believe it , so you do.

No, I believe it because I'm very, very familiar with this subject because of my work and educational background.



Big Blue wasn't supposed to lose to a human either, but it did.

Kasparov lost the year after. He hasn't beaten a top-rank chess playing computer since the late 90s. There was a further 5 years where the best computers and the strongest human grandmasters seemed roughly evenly matched, and then machines have been soundly crushing human grandmasters in chess tournaments since 2004

In 2009 a chess engine running on a mobile phone's processor achieved grandmaster ranking in tournaments against humans.

The human vs computer chess thing is pretty solidly over now. Machines have won.



There is no empirical evidence whatsoever to support the idea that these self-driving vehicles will be flawless and make roads safer.

Yes there is. Their current driving record is better than the human mean *even if you assume Google lied about every single incident*. And they haven't released them to the public yet, they're still testing and improving them.

Feryk
22nd May 15, 10:15 AM
There's no reason to think a machine can do it better either. You just want to believe it , so you do.
Big Blue wasn't supposed to lose to a human either, but it did.
There is no empirical evidence whatsoever to support the idea that these self-driving vehicles will be flawless and make roads safer.

http://techproessentials.com/wp-content/uploads/2015/02/img-video-jeopardy.jpg

Cullion
22nd May 15, 01:50 PM
you joke now, but you'll wish it had happened after the next divorce.

Dr. Socially Liberal Fiscally Conservative Vermin
22nd May 15, 02:20 PM
The RoboBottom is already a reality, Mrs NoB will soon delete him.

Feryk
22nd May 15, 02:30 PM
This thread should be enough proof that if you are worried about losing out to a machine, it's already too late.

Dr. Socially Liberal Fiscally Conservative Vermin
23rd May 15, 05:24 PM
Here are the jobs automation will kill next. (http://fortune.com/2015/05/20/here-are-the-jobs-automation-will-kill-next/)

Short article but pertinent to the thread, think you'll like it OZZ.

Üser Friendly
24th May 15, 01:55 AM
Humans and Autonomy Lab (HAL)

Did they pick the name just to get the 2001 reference in, because it's a pretty stupid name?

The article recommends programming as a way to keep a job, but wont AI take over those functions as well?

OZZ
24th May 15, 06:52 PM
After winning the first time I would never have played Big Blue again, If I was Kasparov.

I can't even beat my fucking laptop at chess.

OZZ
24th May 15, 07:05 PM
Their current driving record is better than the human mean *even if you assume Google lied about every single incident*. And they haven't released them to the public yet, they're still testing and improving them.

So what about transport trucks and heavy machinery ? Has a computer ever driven a rig cross country and been caught in a snowstorm ?
What about dump trucks and cement trucks ? Trucks carrying gasoline tanks ?
Computerized vehicles have yet to do any of this, and have a long way to go and a lot to learn before they are 'better' than humans at all manner of driving.




they're still testing and improving them.
Which means they are not perfect yet.

Cullion
24th May 15, 08:30 PM
So what about transport trucks and heavy machinery ? Has a computer ever driven a rig cross country and been caught in a snowstorm ?
What about dump trucks and cement trucks ? Trucks carrying gasoline tanks ?
Computerized vehicles have yet to do any of this, and have a long way to go and a lot to learn before they are 'better' than humans at all manner of driving.

It is only a matter of time. And not very much time, at that. There's no reason at all to think that these things are easier for a human than they are for a machine.



Which means they are not perfect yet.

Think of all the shitty human drivers you've encountered.

Feryk
25th May 15, 11:54 AM
Humans and Autonomy Lab (HAL)

Did they pick the name just to get the 2001 reference in, because it's a pretty stupid name?

The article recommends programming as a way to keep a job, but wont AI take over those functions as well?

Yep. And faster than other fields, IMO. Longevity in careers will be based on relationships. Until people like their robots and computers better than each other, sales and other relationship/communication based fields will still be human focused.

Üser Friendly
25th May 15, 12:15 PM
Yep. And faster than other fields, IMO. Longevity in careers will be based on relationships. Until people like their robots and computers better than each other, sales and other relationship/communication based fields will still be human focused.

Or until AI can mimic human interaction better than humans

Feryk
25th May 15, 01:16 PM
Or until AI can mimic human interaction better than humans

In all formats. Face to face is still a ways off AFAIK.

Although, we are nothing if not adaptable:

http://www.kurzweilai.net/will-robot-pets-replace-the-real-thing?utm_source=KurzweilAI+Weekly+Newsletter&utm_campaign=cf8c95d01b-UA-946742-1&utm_medium=email&utm_term=0_147a5a48c1-cf8c95d01b-282167421

OZZ
25th May 15, 06:17 PM
Jesus, trucking companies must be salivating at the thought of self-driving transport trucks. No more drivers, no more unions.
Then they can tell the Teamsters to shove it all the way up their asses.

Dr. Socially Liberal Fiscally Conservative Vermin
25th May 15, 06:47 PM
Truck companies are throwing loads of money at this kind of thing.

Cullion
26th May 15, 02:46 AM
Jesus, trucking companies must be salivating at the thought of self-driving transport trucks. No more drivers, no more unions.
Then they can tell the Teamsters to shove it all the way up their asses.

Trucking companies, taxi firms, takeaway delivery etc..

I am genuinely concerned that there is wave of mass unemployment coming.

Üser Friendly
26th May 15, 03:22 AM
Is there any area of life that can not be done better and more efficiently by AI?

With everybody potentially out of a job will we see the rise of Neo-Luddites

Or will we see added value due to be 'human made'...you can taste the difference!

Cullion
26th May 15, 05:24 AM
Man evolves through his tools now. I expect we'll continue the process of merging with our technology. Improvements in AI will improve us.

Our economic system will have to change though, otherwise an increasing portion of the population will simply be seen as 'surplus'.

Dr. Socially Liberal Fiscally Conservative Vermin
26th May 15, 05:54 AM
Is there any area of life that can not be done better and more efficiently by AI?

There is some discussion about creativity, whether AI can genuinely be creative in the way a human can.

Also spontaneous playfulness, which IMO is linked to creativity, isn't something AI is good at.



With everybody potentially out of a job will we see the rise of Neo-Luddites

We're already seeing it just have a read through some of OZZ's posts.



Or will we see added value due to be 'human made'...you can taste the difference!

Yeah I tyhink you're onto sumit there. I think there'll be a market for ' Made by Humans'. There is already a trend back towards individually crafted objects amoungst the middle classes.

Üser Friendly
26th May 15, 07:09 AM
IMO creativity is linked to survival. Working out solutions to new, potentially deadly, situations

If AI has self awareness, and the knowledge that poor performance will eventually lead to that unit's destruction, I imagine AI would become creative in order to stay 'alive'

http://en.wikipedia.org/wiki/Butlerian_Jihad

Cullion
26th May 15, 07:33 AM
AI/Machine learning isn't really sophisticated enough for the kind of creativity we recognize in other humans yet. It's just beginning to master the tasks like 'recognise objects in 3 dimensions', 'follow a route whilst avoiding obstacles and following simple rules about road ettiquette' and 'coordinate limbs on slippery surface' that we expect of simpler animals (although machines can now do these things extraordinarily well).

Things like being good at Chess, which we associate with high intelligence in humans, is mostly down to the very powerful 'brute force' calculating a machine can do, rather than the machine thinking in a fuzzy/'organic' way like it does for the simpler tasks.

Being able to solve more complex problems with objects, tools and simple communication in the way that a crow, dolphin or chimpanzee can is going to come later (almost certainly within our lifetimes though). From there something like human creativity will be within it's grasp, but for now it's best to think of current 'AI' as insect/reptile-like intelligence coupled with powerful machine-based calculation ability and fast reactions.

Dr. Socially Liberal Fiscally Conservative Vermin
26th May 15, 07:38 AM
IMO creativity is linked to survival. Working out solutions to new, potentially deadly, situations

If AI has self awareness, and the knowledge that poor performance will eventually lead to that unit's destruction, I imagine AI would become creative in order to stay 'alive'

http://en.wikipedia.org/wiki/Butlerian_Jihad


Yeah I dont mean creative in the sense of survival but I get your point.

I'm going to be a Mentat when I grow up though.

Üser Friendly
26th May 15, 10:46 AM
Yeah I dont mean creative in the sense of survival but I get your point.

In what way do you mean?

Dr. Socially Liberal Fiscally Conservative Vermin
26th May 15, 10:59 AM
In what way do you mean?

I accept that creativity arose from adaptability in order to survive more efficiently. Also that play arose for the same reasons.

But being playful for its own sake isn't something I can see AI doing soon, if at all. Its one thing to make a game but its another to enjoy playing it.


The aim is to develop an AI system that can intelligently design videogames, as part of an investigation into the ways in which software can design creatively," Cook explains on his blog. He developed his program, known as ANGELINA, with the goal of exploring computational creativity as part of his PhD research. "Can we evolve entire arcade games from nothing?" he asks. "Can we start with literally nothing at all, except a few basic ideas about what a game contains, and ask a computer to design levels, populate them with characters, and wrap it all up in a ruleset that is both challenging and fun?

From: This Eerie Game Was Made by Artificial Intelligence (http://motherboard.vice.com/blog/this-eerie-game-was-made-by-artificial-intelligence)

Üser Friendly
26th May 15, 01:48 PM
What if the benefit mammals gain from play gave benefits similar to developing AI

Whilst the AI may not 'enjoy' playing as we would understand it, there would still be positive feed back due to positive development

After all, what we perceive as enjoyment is merely electrochemical reflex

OZZ
26th May 15, 02:19 PM
I'm no Luddite, only a fool does not take advantage of things like advanced technology.

But the potential for abuse of these technologies by the government is something that worries me, as does the aforementioned wave of unemployable persons as well as a vision of the world where automated 'everything' is the norm.
Large numbers of people living in poverty with no real opportunity in sight means civil unrest, and we all know what that leads to.
I have no children, and the chances I ever will are slim - so I really shouldn't even give a shit, but I do.

Dr. Socially Liberal Fiscally Conservative Vermin
26th May 15, 03:56 PM
What if the benefit mammals gain from play gave benefits similar to developing AI

Whilst the AI may not 'enjoy' playing as we would understand it, there would still be positive feed back due to positive development

Sure there's a parallel.


After all, what we perceive as enjoyment is merely electrochemical reflex

'merely' LOL! You make it sound so trivial & understood, its neither! (that was good comma usage right?)



Large numbers of people living in poverty with no real opportunity in sight means civil unrest, and we all know what that leads to.

x0VDAMv68D0

Feryk
26th May 15, 04:45 PM
People won't necessarily be 'living in poverty' in a world where automation does things better than us.

Imagine the 'poverty' of a world with an unlimited power supply, free versions of every material object you can think of, and transportation that is both accident free and perfectly efficient.

Not a bad baseline. There will obviously be people with a ton of resources available to them. 'Human made' may not be that awesome, but instead of 'free table #5' from Walmart, a hand crafted, hand carved table made by Bob would be expensive.

And all such goods would be. The ultimate expression of human endeavor is art. So we would all become artists in some way.

Dr. Socially Liberal Fiscally Conservative Vermin
26th May 15, 04:48 PM
WORD! BROTHER FERKY!!!

Feryk
26th May 15, 04:49 PM
And for those who want to know: the color of the sky in my world is cerulean.

Cullion
26th May 15, 04:53 PM
I accept that creativity arose from adaptability in order to survive more efficiently. Also that play arose for the same reasons.

But being playful for its own sake isn't something I can see AI doing soon, if at all. Its one thing to make a game but its another to enjoy playing it.

Define 'enjoy'

Feryk
26th May 15, 04:55 PM
Endorphin release?

Dr. Socially Liberal Fiscally Conservative Vermin
26th May 15, 04:56 PM
Yeah tricky, like Doofa has already alluded to, do think a machine can feel?

Feryk
26th May 15, 04:59 PM
Not yet, but soon.

Very soon.

http://motherboard.vice.com/blog/emotionally-intelligent-machines-are-closer-than-ever

Dr. Socially Liberal Fiscally Conservative Vermin
26th May 15, 05:04 PM
So is the proposal that we are just meat puppets?

Feryk
26th May 15, 05:09 PM
Do you feel like a meat puppet?

Why would you think that?

Cullion
26th May 15, 05:12 PM
Yeah tricky, like Doofa has already alluded to, do think a machine can feel?

As much as an insect or simple reptile can, at the moment.

Dr. Socially Liberal Fiscally Conservative Vermin
26th May 15, 05:19 PM
Do you feel like a meat puppet?

Why would you think that?

Well if any function we can perform is ultimately reproducible via man made puppets (AI, robots etc) then why are we any different?

Dr. Socially Liberal Fiscally Conservative Vermin
26th May 15, 05:20 PM
As much as an insect or simple reptile can, at the moment.

I'm not sold on that as it stands. Reptiles are are capable of quite complex emotional responses that I have as yet to see from AI.

Cullion
26th May 15, 05:31 PM
I think Maslow's hierarchy of needs is relevant here.

Higher forms of behaviour are rooted in the layer below. Emotional responses are rooted in basic needs of an organism. Hunger, fear, lust etc.. Machines are usually built with different fundamental drives, so no matter how sophisticated their mental processes, I think you would always find their behaviour somehow cold and alien unless they were deliberately built to mimic human/mammalian drives at a very basic level.

The insect/reptile-like machines we have today have not been built with those basic animal-mimicking drives.

Üser Friendly
27th May 15, 03:02 AM
merely' LOL! You make it sound so trivial & understood, its neither! (that was good comma usage right?)

Are you positing the existence of a soul?

can you write code for that?

Pie of Hate
27th May 15, 03:10 AM
People won't necessarily be 'living in poverty' in a world where automation does things better than us.

Imagine the 'poverty' of a world with an unlimited power supply, free versions of every material object you can think of, and transportation that is both accident free and perfectly efficient.

Not a bad baseline. There will obviously be people with a ton of resources available to them. 'Human made' may not be that awesome, but instead of 'free table #5' from Walmart, a hand crafted, hand carved table made by Bob would be expensive.

And all such goods would be. The ultimate expression of human endeavor is art. So we would all become artists in some way.

This is just one step closer to Star Trek!

Feryk
27th May 15, 10:40 AM
Yes, it is. And all the problems that Next Gen had with their computers, too.

Pie of Hate
28th May 15, 03:01 AM
Yes, it is. And all the problems that Next Gen had with their computers, too.

They only had problems because they didn't have an Indian call centre to say "have you tried turning it off, then restarting?"

Üser Friendly
28th May 15, 03:04 AM
They outsourced all the call centers to the Ferengi

Spade: The Real Snake
28th May 15, 10:07 AM
Hawking better hope the machines don't become sentient, otherwise his ass will be the first dumped over the White Cliffs of Dover.
And without his crip-puter, nobody will hear him scream.

OZZ
28th May 15, 01:33 PM
Imagine the.... world with an unlimited power supply, free versions of every material object you can think of, and transportation that is both accident free and perfectly efficient.

Not a bad baseline. There will obviously be people with a ton of resources available to them. 'Human made' may not be that awesome, but instead of 'free table #5' from Walmart, a hand crafted, hand carved table made by Bob would be expensive.

And all such goods would be. The ultimate expression of human endeavor is art. So we would all become artists in some way.

I'm not so sure about the 'free versions of every material object you can think of ' thing, but whatever..
Interesting thing about it though; I took a Sociology of Work and Industry course waaay back during 1st yr. undergrad studies with a very clever fellow named Pete Stemp.
Another non -PhD'er who had more to contribute to teaching than the majority of tenured profs I've met.
Anyways, I recall his prophetic vision of the future being quite like the one you describe above, and in retrospect it seems that he was very on track with a lot of his notions about the direction that Western Society is headed as far as work and industry is concerned.
A good number of people working from their homes doing creative, mentally stimulating work with low stress levels. Less pollution and lots of renewable, non-toxic energy to go around.
I should pull out those notes and take a look at them, it feels like a lifetime ago I was listening to that guy.
I really enjoyed that class, even though (for the most part) I loathe sociology...

Feryk
28th May 15, 04:27 PM
They only had problems because they didn't have an Indian call centre to say "have you tried turning it off, then restarting?"
http://rlv.zcache.com/have_you_tried_turning_it_off_and_on_again_mug-r536dec88b320466daedda28c9e0319b3_x7jgm_8byvr_324. jpg

I love the IT Crowd. Deliciously stupid fun.

Feryk
28th May 15, 04:28 PM
I'm not so sure about the 'free versions of every material object you can think of ' thing, but whatever..
Interesting thing about it though; I took a Sociology of Work and Industry course waaay back during 1st yr. undergrad studies with a very clever fellow named Pete Stemp.
Another non -PhD'er who had more to contribute to teaching than the majority of tenured profs I've met.
Anyways, I recall his prophetic vision of the future being quite like the one you describe above, and in retrospect it seems that he was very on track with a lot of his notions about the direction that Western Society is headed as far as work and industry is concerned.
A good number of people working from their homes doing creative, mentally stimulating work with low stress levels. Less pollution and lots of renewable, non-toxic energy to go around.
I should pull out those notes and take a look at them, it feels like a lifetime ago I was listening to that guy.
I really enjoyed that class, even though (for the most part) I loathe sociology...

I'm more interested in your philosophical take on the matter in the Build a God 1.0 thread.

Dr. Socially Liberal Fiscally Conservative Vermin
29th May 15, 01:39 PM
I know a lot of humans who havent got this level coordination...

_luhn7TLfWU

Üser Friendly
29th May 15, 02:07 PM
That's fantastic

Fit it with machine guns and we're good to go

Feryk
29th May 15, 04:58 PM
Make it run at 60 mph, and add a counterweighted tail it can throw out to change direction and I'll be impressed.

Dr. Socially Liberal Fiscally Conservative Vermin
30th May 15, 05:16 AM
I couldn't do that.

Feryk
2nd June 15, 11:37 AM
Nope. Not at all.

Dr. Socially Liberal Fiscally Conservative Vermin
2nd June 15, 12:23 PM
I knew it!

BOW DOWN TO YOUR ROBOTIC SUPERIORS!!!!!!

Feryk
2nd June 15, 12:52 PM
One day maybe, but laser cheetahbot is not it.

Üser Friendly
2nd June 15, 01:14 PM
What?

Feryk can't run?

Feryk
2nd June 15, 01:15 PM
Not like a cheetah, and neither can you.

Üser Friendly
2nd June 15, 01:24 PM
I am SVELTE! like a cheetah

Üser Friendly
19th June 15, 12:40 PM
http://www.dailymail.co.uk/tvshowbiz/article-3130153/They-won-t-Arnold-Schwarzenegger-67-pranks-unwitting-fans-Madame-Tussauds-posing-Terminator-waxwork-coming-life.html

NoBowie
24th March 16, 02:46 PM
This seems to align with Hawking's assertion:

Recently, Microsoft put an AI experiment onto Twitter, naming it "Tay". The bot was built to be fully aware of the latest adolescent fixations (e.g. celebrities and similar), and to interact like a typical teen girl.

In less than 24 hours, it inexplicably became a neo-nazi sex robot with daddy issues. Sample tweets from it proclaimed that "Hitler did nothing wrong!", then went on to blame former President Bush for 9/11, stated that "donald trump is the only hope we've got", and other similar instances.

As the hours passed, it all went downhill from there, eventually spewing racial slurs and profanity, demanding sex, and calling everyone "daddy". The bot was quickly removed once Microsoft discovered the trouble.

-------------

"Fuck my robot pussy" This is making me laugh so hard. 4chan was responsible for this mayhem, as usual. Watched it unfold on 4chan myself.

http://i.imgur.com/Wb3QZ2j.jpg

http://www.telegraph.co.uk/technology/2016/03/24/microsofts-teen-girl-ai-turns-into-a-hitler-loving-sex-robot-wit/




(http://www.telegraph.co.uk/technology/2016/03/24/microsofts-teen-girl-ai-turns-into-a-hitler-loving-sex-robot-wit/)

Syntactical Disruptorize
26th March 16, 10:36 AM
This seems to align with Hawking's assertion:
♫ Best thing ever best thing ever this is the best thing ever♪

Feryk
28th March 16, 10:37 AM
So apparently the AI was easily influenced by assholes. Surprise.

NoBowie
29th March 16, 12:44 PM
Canada wants to open the door to skynet:

A QUARTER OF CANADIAN ADULTS BELIEVE A COMPUTER PROGRAM WOULD BE MORE TRUSTWORTHY AND ETHICAL THAN THEIR BOSS

http://www.intensions.co/news/2016/3/29/intensions-future-of-work

Feryk
29th March 16, 12:53 PM
They just haven't figured out how good AI can become at lying.

NoBowie
11th July 16, 03:59 PM
Don't piss off your doc-bot!

"robot performed the world's first autonomous soft-tissue surgery without any assistance from doctors, outperformed human doctors by every metric except for speed"

http://www.popularmechanics.com/science/health/a20718/first-autonomous-soft-tissue-surgery/

NoBowie
6th September 16, 10:39 AM
First AI created movie trailer, Watson analyzed movie trailers and then chose the scenes:

rqmHSR0bFU8

http://www.popsci.com/ibm-watson-created-first-ever-ai-made-trailer-for-morgan

"IBM used machine learning and experimental Watson APIs, parsing out the trailers of 100 horror movies. It did visual, audio, and composition analysis of individual scenes, finding what makes each moment eerie, how the score and actors' tone of voice changed the mood--framing and lighting came together to make a complete trailer. Watson was then fed the full film, and it chose scenes for the trailer. A human--in this case, the "resident IBM filmmaker"--still needed to step in to edit for creativity. Even so, a process that would normally take weeks was reduced to hours."

-------------------

What if AI replaces artists before it replaces other vocations?