Post: Artificial Super Intelligence
Options
12-04-2015, 05:34 AM
#1
partypoker15
Call me G
Lets say, hypothetically, that the government managed to build, or buy, an ASI that could do exactly this. Watch anything and everything. Determine premeditated acts of violence, terrorism or otherwise, and alert authorities to the crime. The catch is the system is a black box, meaning you can't see inside. The government could not see any of the information gathered by this ASI, and could only receive the social security number of the person. The catch is, if the crime is not an act of terror, the government would have no way of knowing if the person was a victim, or perpetrator. Would you have a problem with this? If it was a machine that watched over you, not a human being, would you be okay with mass surveillance and information processing on this scale? Would you feel it an invasion of privacy?
Personally, I would have no problem with an ASI such as this. If you can't get any information out of it, other than the social security number of a perpetrator or possible victim, and it could help stop premeditated homicide and planned terrorist attacks, I would gladly give up my small bit of privacy. On one condition, the ASI will only intervene if the situation is seen as lethal. Meaning if no one would die from the crimes that are being planned, the ASI will ignore the crime. I do not want to live in a world where you can be prosecuted for thinking about committing a crime, otherwise I'd already be in jail. There is a flaw in this, crimes of opportunity, lethal or not, cannot be seen. If a person saw someone they wanted dead out walking on the street, there is no algorithm that can predict he will kill. The algorithms can only analyze and detect patterns, in other words, it can only see premeditated acts.
So what is your opinion? I honestly think that even though my idea for this discussion came from a TV show, it could still end up being a good discussion.
12-05-2015, 06:53 PM
#2
Originally posted by partypoker15
So a little bit of background before I start. Over the past month I have been binge watching Person of Interest, a CBS crime drama that is based around the idea that the government is in possession of an Artificial Super Intelligence (ASI) that can see anything and everything connected to a network. Phones, emails, texts, security cameras, you get the point. It uses this information to determine premeditated crimes. Premeditated acts of terror are considered "relevant" as this ASI was designed with a goal of stopping acts of terror on U.S. soil after 9/11, and the government receives the social security number or the terrorist. After observation, the terrorist is killed by government agents. Any other premeditated act of violence (usually a murder) is considered "irrelevant" and the government never receives the information. I don't want to explain the entire plot of the show, so that is all the background I need to start the discussion.
Lets say, hypothetically, that the government managed to build, or buy, an ASI that could do exactly this. Watch anything and everything. Determine premeditated acts of violence, terrorism or otherwise, and alert authorities to the crime. The catch is the system is a black box, meaning you can't see inside. The government could not see any of the information gathered by this ASI, and could only receive the social security number of the person. The catch is, if the crime is not an act of terror, the government would have no way of knowing if the person was a victim, or perpetrator. Would you have a problem with this? If it was a machine that watched over you, not a human being, would you be okay with mass surveillance and information processing on this scale? Would you feel it an invasion of privacy?
Personally, I would have no problem with an ASI such as this. If you can't get any information out of it, other than the social security number of a perpetrator or possible victim, and it could help stop premeditated homicide and planned terrorist attacks, I would gladly give up my small bit of privacy. On one condition, the ASI will only intervene if the situation is seen as lethal. Meaning if no one would die from the crimes that are being planned, the ASI will ignore the crime. I do not want to live in a world where you can be prosecuted for thinking about committing a crime, otherwise I'd already be in jail. There is a flaw in this, crimes of opportunity, lethal or not, cannot be seen. If a person saw someone they wanted dead out walking on the street, there is no algorithm that can predict he will kill. The algorithms can only analyze and detect patterns, in other words, it can only see premeditated acts.
So what is your opinion? I honestly think that even though my idea for this discussion came from a TV show, it could still end up being a good discussion.
Lets say, hypothetically, that the government managed to build, or buy, an ASI that could do exactly this. Watch anything and everything. Determine premeditated acts of violence, terrorism or otherwise, and alert authorities to the crime. The catch is the system is a black box, meaning you can't see inside. The government could not see any of the information gathered by this ASI, and could only receive the social security number of the person. The catch is, if the crime is not an act of terror, the government would have no way of knowing if the person was a victim, or perpetrator. Would you have a problem with this? If it was a machine that watched over you, not a human being, would you be okay with mass surveillance and information processing on this scale? Would you feel it an invasion of privacy?
Personally, I would have no problem with an ASI such as this. If you can't get any information out of it, other than the social security number of a perpetrator or possible victim, and it could help stop premeditated homicide and planned terrorist attacks, I would gladly give up my small bit of privacy. On one condition, the ASI will only intervene if the situation is seen as lethal. Meaning if no one would die from the crimes that are being planned, the ASI will ignore the crime. I do not want to live in a world where you can be prosecuted for thinking about committing a crime, otherwise I'd already be in jail. There is a flaw in this, crimes of opportunity, lethal or not, cannot be seen. If a person saw someone they wanted dead out walking on the street, there is no algorithm that can predict he will kill. The algorithms can only analyze and detect patterns, in other words, it can only see premeditated acts.
So what is your opinion? I honestly think that even though my idea for this discussion came from a TV show, it could still end up being a good discussion.
Moved to computers for further discussion bruv.
12-08-2015, 12:01 AM
#3
SoulAscension
Haxor!
12-08-2015, 01:24 PM
#4
Bch
Guest
Originally posted by partypoker15
So a little bit of background before I start. Over the past month I have been binge watching Person of Interest, a CBS crime drama that is based around the idea that the government is in possession of an Artificial Super Intelligence (ASI) that can see anything and everything connected to a network. Phones, emails, texts, security cameras, you get the point. It uses this information to determine premeditated crimes. Premeditated acts of terror are considered "relevant" as this ASI was designed with a goal of stopping acts of terror on U.S. soil after 9/11, and the government receives the social security number or the terrorist. After observation, the terrorist is killed by government agents. Any other premeditated act of violence (usually a murder) is considered "irrelevant" and the government never receives the information. I don't want to explain the entire plot of the show, so that is all the background I need to start the discussion.
Lets say, hypothetically, that the government managed to build, or buy, an ASI that could do exactly this. Watch anything and everything. Determine premeditated acts of violence, terrorism or otherwise, and alert authorities to the crime. The catch is the system is a black box, meaning you can't see inside. The government could not see any of the information gathered by this ASI, and could only receive the social security number of the person. The catch is, if the crime is not an act of terror, the government would have no way of knowing if the person was a victim, or perpetrator. Would you have a problem with this? If it was a machine that watched over you, not a human being, would you be okay with mass surveillance and information processing on this scale? Would you feel it an invasion of privacy?
Personally, I would have no problem with an ASI such as this. If you can't get any information out of it, other than the social security number of a perpetrator or possible victim, and it could help stop premeditated homicide and planned terrorist attacks, I would gladly give up my small bit of privacy. On one condition, the ASI will only intervene if the situation is seen as lethal. Meaning if no one would die from the crimes that are being planned, the ASI will ignore the crime. I do not want to live in a world where you can be prosecuted for thinking about committing a crime, otherwise I'd already be in jail. There is a flaw in this, crimes of opportunity, lethal or not, cannot be seen. If a person saw someone they wanted dead out walking on the street, there is no algorithm that can predict he will kill. The algorithms can only analyze and detect patterns, in other words, it can only see premeditated acts.
So what is your opinion? I honestly think that even though my idea for this discussion came from a TV show, it could still end up being a good discussion.
Lets say, hypothetically, that the government managed to build, or buy, an ASI that could do exactly this. Watch anything and everything. Determine premeditated acts of violence, terrorism or otherwise, and alert authorities to the crime. The catch is the system is a black box, meaning you can't see inside. The government could not see any of the information gathered by this ASI, and could only receive the social security number of the person. The catch is, if the crime is not an act of terror, the government would have no way of knowing if the person was a victim, or perpetrator. Would you have a problem with this? If it was a machine that watched over you, not a human being, would you be okay with mass surveillance and information processing on this scale? Would you feel it an invasion of privacy?
Personally, I would have no problem with an ASI such as this. If you can't get any information out of it, other than the social security number of a perpetrator or possible victim, and it could help stop premeditated homicide and planned terrorist attacks, I would gladly give up my small bit of privacy. On one condition, the ASI will only intervene if the situation is seen as lethal. Meaning if no one would die from the crimes that are being planned, the ASI will ignore the crime. I do not want to live in a world where you can be prosecuted for thinking about committing a crime, otherwise I'd already be in jail. There is a flaw in this, crimes of opportunity, lethal or not, cannot be seen. If a person saw someone they wanted dead out walking on the street, there is no algorithm that can predict he will kill. The algorithms can only analyze and detect patterns, in other words, it can only see premeditated acts.
So what is your opinion? I honestly think that even though my idea for this discussion came from a TV show, it could still end up being a good discussion.
If this was done secretly and undetected, i think it would be a great way to go against massive organised crime, not so much your petty shop lifter stealing a bag of haribos, but the large scale drug operations, the large scale terrorist operations etc. With the rapid growth in technology, im sure this won't be too far away, it's totally possible but if it was done with everyones knowledge, people wouldn't agree, and large organised crimes will have the money to buy large separate networks with special customised phones etc so i think it would deffo need to be done secretly.
12-08-2015, 04:44 PM
#5
partypoker15
Call me G
Originally posted by SoulAscension
A rule of thumb for me is to never trust artificial 'intelligence' with these sorts of things. And I just wouldn't trust the people 'regulating' this black box
That is the interesting thing about it being a black box. There would be no way for the people "regulating" it to get any information out of it. Only a number of a perpetrator or victim.
Copyright © 2024, NextGenUpdate.
All Rights Reserved.
