A greater intelligence may prefer internally inconsistent rules and use complex heuristics to choose between them. To us, the results of that process may seem so unfathomably complex that it looks random.
I hope you are right. On the other hand, they may decide that we will live and die by the rules that we created in some sort of ethical calculus involving self-determination.
The robots might be cheaper today. Especially if it is just a webcam, microphone, motorized wheels, battery and the cheapest wireless networking computer being operated by a server in a more secure area of the store.
I imagine a single guard monitoring feeds from ten of these roaming around a wal-mart. The guard doesn't even have to be in the wal-mart. Throw a blanket over it and the guard knows something is wrong and calls physical security.
Throw in some advanced mapping that can compare expected camera images to actual images and use the human to examine positive signals and dismiss the false positives. You might be able to up that human monitor to monitor 100 to 1000 units depending on the false positive rate and the rate of actual incidents.
99% of the time, nothing is happening during security guard duty.
In this case, all lower paid employees will be outsourced to a contracting company that may or may not be a fully owned subsidiary of the original employer.
This is just another example of the government cutting funding for the arts. Sure, it may be security theatre but these days that is the only kind of theatre I see to have time for.
Maybe we can get the National Endowment for the Arts to pick up the slack. Or they could move to an NPR model and hold pledge drives.
He's talking about glasses-free holographic 3D that perfectly reconstructs the light field you would see if you were looking at the real object. With that you get motion parallax, stereo perception and the ability to focus.
That is easy. Break the task down into smaller tasks that you can estimate. If you can't estimate a broken down task, break it down more (think "3 hours"). Then add up all the bits.
1. Figuring out the right problem to solve. Or the best problem (which problem will have the most impact when solved). 2. Explaining why something took longer than I expected it to. 3. Estimating the time to fix some random bug before I understand the bug. 4. Explaining why the work I've done over the past year is more important/difficult/better than the work done by a co-worker over the past year. 5. Balancing staying focused on task versus expanding the breadth of my knowledge about the system I'm working on (I naturally bias towards the former). 6. Understanding the forest when I spend all day looking at individual trees (related to 5). 7. Knowing when to ask a co-worker a question to get past an obstacle in five minutes versus spending an hour figuring it out on my own.
Most of the things the article describe sound like the easy parts.
Most of my challenges seem to revolve around communication.
The Kinect gets 3D information. They can take the polygons mapped with the video image and apply small rotation from the picture of the other person onthe screen to the camera. You'll get some artifacts where you are exposing areas that the camera can't see, but maybe those can be filled in with surrounding pixels so you don't notice so much.
My only complaint is that I am skeptical a planet sized brain would be that aggressive. I'd rather think it would consider humans part of its own 'biology' in much the same way that we consider mitochondria and e. coli part of our biology.
We have no problem wiping out e. coli when it gets out of line but we would never eradicate it completely.
Perhaps the larger the team, the longer the short term is.
The average lifespan of a human being should then predict how selfish we are as a species and how large our social organizations (businesses, governments, churches, cities, etc) can be before corruption sets in.
It also implies the only way to really get people to cooperate is through life extension technologies.
Slashdot Mirror
A greater intelligence may prefer internally inconsistent rules and use complex heuristics to choose between them. To us, the results of that process may seem so unfathomably complex that it looks random.
I enjoyed this story which is related: http://www.ssec.wisc.edu/~billh/g/mcnrsts.html
I hope you are right. On the other hand, they may decide that we will live and die by the rules that we created in some sort of ethical calculus involving self-determination.
Bravo!
But the AI's will be able to move. They'll be able to move very quickly, indeed.
The robots might be cheaper today. Especially if it is just a webcam, microphone, motorized wheels, battery and the cheapest wireless networking computer being operated by a server in a more secure area of the store.
I imagine a single guard monitoring feeds from ten of these roaming around a wal-mart. The guard doesn't even have to be in the wal-mart. Throw a blanket over it and the guard knows something is wrong and calls physical security.
Throw in some advanced mapping that can compare expected camera images to actual images and use the human to examine positive signals and dismiss the false positives. You might be able to up that human monitor to monitor 100 to 1000 units depending on the false positive rate and the rate of actual incidents.
99% of the time, nothing is happening during security guard duty.
OMG. I didn't realize it wasn't K9 until you pointed it out.
I'm happy that a few versions from now it might be vaguely dog shaped.
And maybe it will have a helpful Death Ray in its nose.
http://www.youtube.com/watch?v=8HgejSCHRi8
And confusing the Doctor by naming it K-9.
Fool! Don't you know that Daleks can levitate?
This decision will also be used precedence by the machines to decide how humans should be treated post-singularity. Choose wisely.
Yep.
In this case, all lower paid employees will be outsourced to a contracting company that may or may not be a fully owned subsidiary of the original employer.
I think I may love you.
The previous sentence may contain a typo.
I have some stand-up open mics on youtube.
https://www.youtube.com/channel/UCoJjJOzT5vi4FqOOQXaUiNw
This is just another example of the government cutting funding for the arts. Sure, it may be security theatre but these days that is the only kind of theatre I see to have time for.
Maybe we can get the National Endowment for the Arts to pick up the slack. Or they could move to an NPR model and hold pledge drives.
They make money the old fashion way: Volume.
He's talking about glasses-free holographic 3D that perfectly reconstructs the light field you would see if you were looking at the real object. With that you get motion parallax, stereo perception and the ability to focus.
Create milestones. Only change the spec like that between milestones and get buy-off on the design changes and impact to the schedule.
That is easy. Break the task down into smaller tasks that you can estimate. If you can't estimate a broken down task, break it down more (think "3 hours"). Then add up all the bits.
1. Figuring out the right problem to solve. Or the best problem (which problem will have the most impact when solved).
2. Explaining why something took longer than I expected it to.
3. Estimating the time to fix some random bug before I understand the bug.
4. Explaining why the work I've done over the past year is more important/difficult/better than the work done by a co-worker over the past year.
5. Balancing staying focused on task versus expanding the breadth of my knowledge about the system I'm working on (I naturally bias towards the former).
6. Understanding the forest when I spend all day looking at individual trees (related to 5).
7. Knowing when to ask a co-worker a question to get past an obstacle in five minutes versus spending an hour figuring it out on my own.
Most of the things the article describe sound like the easy parts.
Most of my challenges seem to revolve around communication.
Psychohistory kind of depends on most people nothing thinking of Psychohistory.
The Kinect gets 3D information. They can take the polygons mapped with the video image and apply small rotation from the picture of the other person onthe screen to the camera. You'll get some artifacts where you are exposing areas that the camera can't see, but maybe those can be filled in with surrounding pixels so you don't notice so much.
I read "followed by" too, but I just assumed the slashdot summary was written badly. Even though I was wrong, I'm sticking to that theory.
I think most Americans agree that our health care system is broken, but there is a lot of disagreement on the way to fix it.
My only complaint is that I am skeptical a planet sized brain would be that aggressive. I'd rather think it would consider humans part of its own 'biology' in much the same way that we consider mitochondria and e. coli part of our biology.
We have no problem wiping out e. coli when it gets out of line but we would never eradicate it completely.
the Traveller?
I thought this was brilliant.
Perhaps the larger the team, the longer the short term is.
The average lifespan of a human being should then predict how selfish we are as a species and how large our social organizations (businesses, governments, churches, cities, etc) can be before corruption sets in.
It also implies the only way to really get people to cooperate is through life extension technologies.