Opinion: How artificial intelligence can predict mass shootings

  • Thread starter Thread starter TK421
  • Start date Start date
Status
Not open for further replies.
I’m hopeful that big data can warn us of impending terror attacks. I’m also skeptical how it will be used. I saw Minority Report! The tech itself isn’t evil…it’s a tool and it’s all in how it’s used. And please, with caution.

Another question is does this violate our right to privacy? I don’t think combing the data is itself a violation but the next steps taken might very well be a violation.
 
Another question is does this violate our right to privacy? I don’t think combing the data is itself a violation but the next steps taken might very well be a violation.
Our privacy laws as written don’t really cover inferences from big data. All of the little bits of data individually are of little meaning. The inferences that one can make from them depend in part on computational capabilities. I think it would be hard to make legislation for it, since these capabilities change all the time.

Making rules for machine intelligence can be challenging, as any rule has to be in the form of a computational constrain. For many of the things that people concern themselves with don’t convert to computations very well.
I’m hopeful that big data can warn us of impending terror attacks.
Would be nice. I think they will also have to give considerations on fairly dealing with false positives. There have been a couple of people in the past years that have found themselves in unfortunate situations because of ML. In the area of face recognition the ML was typically made with a low number of samples (if any) from people with dark skin or from women. Thus many face recognition algorithms have higher error rates if someone has dark skin or is a woman. There have been a few people flagged as bing criminals from false positives and found themselves in a jail until the machine’s mistakes were sorted out.
 
Another question is does this violate our right to privacy? I don’t think combing the data is itself a violation but the next steps taken might very well be a violation.
Well, when you post something on the internet, it is there for others to view.

I’m also a bit concerned of it being used overly aggressively and there will need to be some humans reviewing the data, but AI can still cut out a lot of the work.
 
Last edited:
I’m also a bit concerned of it being used overly aggressively and there will need to be some humans reviewing the data, but AI can still cut out a lot of the work.
The algorithms will probably rather primitive when it’s first started but will improve with time…a reason not to rush the program! There are already data mining companies that are uploading every picture posted in Facebook. As the public becomes more aware…if they ever do…on how big a trail they leave just on Facebook alone, perhaps people will also start demanding laws regulating what that data can be used for. Right now, it’s mostly to serve us ads…and it’s mostly pretty awful at it! Buy something on Amazon and you get blasted for weeks with the same item…umm, I just bought it!

There have been issues with the Ring doorbell having law enforcement subpoenaing the camera roll. Blacks and women misidentified as @ThinkingSapien said and other problems and what many feel is a violation of privacy. It also moves and develops so fast…much faster than congress can keep up with…if they even understand it.

Now, there are battles going on with encryption…government and law wanting back door access and the tech world denying them any right to our own data. It’ll probably get much uglier before we find our footing and abuses brought under control. It’s amazing how little data is needed to ID a person!
 
Now, there are battles going on with encryption…government and law wanting back door access and the tech world denying them any right to our own data.
I think the government request betray their level of understanding of how encryption works. Encryption, while not impossible to defeat, is computationally expensive to defeat. Further, part of the success of encryption is based on limited accessibility to the keys. The foundations of the encryption that is used today is based on research that happened during World War II. In P2P encryption the middle-man never knows the encryption keys. There is no back door to that lack of knowledge. Any attempt to pass on that encryption key opens vulnerabilities to a lot of parties that to which consent was not given. Not just the service provider or the government.

Additionally, such demands only open a market for message providers with footprints outside of the USA that have no obligations to the back-door requirements. It makes the law-abiding citizens less secure while those trying to hide activity still have access to encryption that is computationally expensive to defeat.
 
The ability of legislators to keep up with the tech world is one of the biggest obstacles, plus the way data is used can be so nuanced that legislators can pass a certain regulation and then companies can tip toe around it with relative ease and have a team of lawyers to defend it.

I remember Andrew Yang talking about how Jeff Bezos is in some way more powerful than the POTUS, which I still think is an exaggeration but if the tech world can’t be properly regulated than it might come true.
 
Last edited:
Excellent post! 👏 👏 👏 👏 👏

The failure of our own government agencies to keep their hacking tools secure is proof that having any agency have a back door leads to nefarious people getting it, too!
 
The failure of our own government agencies
I think there is a failure of our governments understanding of technology in general. Many of our representatives have a legal background. Few of them have backgrounds in anything else. As a Software Engineer (by occupation and degree) I watch the technology related hearings that the representatives hold. Unfortunately they don’t understand what they are seeking to legislate and have to work from metaphors. Metaphors of how something works is not an explanation of how it works. So I feel like the decisions that come out of the hearing are at a very low level of understanding (if I can call what they achieve “understanding”).
to keep their hacking tools secure
True. Leak of government security tools have made us less secure.

Those TSA approved locks…A TSA official did an interview in which he let someone take a picture of the TSA keyring of TSA keys. Someone produced 3D print plans for the keys from the photograph. Thus anyone that knows how to use Google and knows how to order a 3D print (assuming they don’t have a 3D printer) can have their own copy of keys for the TSA approved locks!
have a back door leads to nefarious people getting it, too!
Yep! Every point of access is a point of access that a nefarious character could potentially use.
 
Seems messed up to me. Not sure I want organisations to be amassing data on me as a means to predict what crimes I might commit in the future. I know it’s how things are going anyway, but how anyone can feel secure in knowing this is happening I don’t understand.
 
@Regular_Atheist

I guess my way of looking at it is that we can get started now and then governments and society in general can start reacting and regulating as needed. Ignoring it can’t make it go away and using AI to predict actions is only going to get better.
 
Last edited:
Seems messed up to me. Not sure I want organisations to be amassing data on me as a means to predict what crimes I might commit in the future.
Also, keep in mind that the data collection might not limit their inference capabilities to only crime. There are other aspects of personal life that they might be able to figure out.
I guess my way of looking at it is that we can get started now and then governments and society in general can start reacting and regulating as needed.
Taking a reactive approach instead of a proactive approach on regulations could result in a lot of I justices occurring and it being too late to effectively fix the problems.

A proactive strategy would include having representatives get a basic understanding of AI, putting them in a better position to make laws concerning it.
 
Last edited:
Well, best of luck to us. Representatives are pretty much all lawyers with zero background in tech.
 
Last edited:
A lot of times, these are just data crawlers that utilize anything that can be searched for in google search. If they have access to user data, private messages, phone numbers, text messages, then that’s a different story…but usually it’s bot crawlers taking advantage of forum/website search engine optimization…Then a program interprets the data.
 
Status
Not open for further replies.
Back
Top