View previous topic | View next topic

Internet privacy?

Page 3 of 4
Goto page Previous  1, 2, 3, 4  Next

CB27
1379265.  Mon Apr 12, 2021 2:06 pm Reply with quote

Going back to the original rant :)

It can easily be argued that Facebook, Twitter, TikTok et al can do a lot more, I wouldn't argue with that, but I would argue against the idea they're doing nothing at all.

As I understand it, there is a two pronged approach at play, and I've seen both in action.

The first is an AI to capture certain language (not so much words, but phrases), and these are automatically flagged.

The second is the facility for users to flag posts they deem offensive or inappropriate.

In each case there will then be moderators, and while social media companies don't usually employ these direct (for various reasons, but mostly to do with tax because of business status, etc), they do actually employ thousands of moderators through contractors. I don't know what the figure is in the UK, or whether those employed abroad also moderate UK pages and users, but I read during the early days of the pandemic Facebook had 15k moderators in the US alone. I think these businesses are doing well enough that they can afford to employ more moderators, but these are also not insignificant numbers.

The biggest problem I find with online abuse is that people can get away with doing anonymously, and I think that's probably one of the biggest drivers for how big the problem has become. The problem starts with the fact that accounts are created by having an email, and there is no control over these.

Ironically, when Facebook first started you needed to have a university email to create an account - I know because I had to get a friend to create an account for me back in the day. Perhaps Governments need to consider legislating that social media accounts can only be created using registered emails. It doesn't stop people using different emails for communications or logging into other types of platforms, nor will it give Governments any access to social media account details, but if someone is suspended or banned, it can stop them from creating other accounts in other names. Similarly it can slow down the use of bots and the spread of misinformation.

 
CB27
1379266.  Mon Apr 12, 2021 2:11 pm Reply with quote

cornixt wrote:
We had those in the earlier days of the internet, and the fact that we don't any more is a demonstration of how well they worked.

I remember the days of the usenet groups, it was scary how you could accidentally stumble onto things that would probably warrant a knock on your door now if you were to google them.

 
Brock
1379270.  Mon Apr 12, 2021 2:22 pm Reply with quote

CB27 wrote:
cornixt wrote:
We had those in the earlier days of the internet, and the fact that we don't any more is a demonstration of how well they worked.

I remember the days of the usenet groups, it was scary how you could accidentally stumble onto things that would probably warrant a knock on your door now if you were to google them.


Usenet isn't dead - it's still going. A lot of groups are filled up with spam, but there are one or two places where it's still possible to have an intelligent discussion. Much quieter than it used to be though.

 
jaygeemack
1379386.  Wed Apr 14, 2021 3:02 am Reply with quote

In future, people will be desperate for 15 minutes of anonymity.

 
crissdee
1379393.  Wed Apr 14, 2021 4:26 am Reply with quote

Nice one sir!

 
Prof Wind Up Merchant
1379399.  Wed Apr 14, 2021 5:44 am Reply with quote

CB27 wrote:
Going back to the original rant :)

It can easily be argued that Facebook, Twitter, TikTok et al can do a lot more, I wouldn't argue with that, but I would argue against the idea they're doing nothing at all.

As I understand it, there is a two pronged approach at play, and I've seen both in action.

The first is an AI to capture certain language (not so much words, but phrases), and these are automatically flagged.

The second is the facility for users to flag posts they deem offensive or inappropriate.

In each case there will then be moderators, and while social media companies don't usually employ these direct (for various reasons, but mostly to do with tax because of business status, etc), they do actually employ thousands of moderators through contractors. I don't know what the figure is in the UK, or whether those employed abroad also moderate UK pages and users, but I read during the early days of the pandemic Facebook had 15k moderators in the US alone. I think these businesses are doing well enough that they can afford to employ more moderators, but these are also not insignificant numbers.

The biggest problem I find with online abuse is that people can get away with doing anonymously, and I think that's probably one of the biggest drivers for how big the problem has become. The problem starts with the fact that accounts are created by having an email, and there is no control over these.

Ironically, when Facebook first started you needed to have a university email to create an account - I know because I had to get a friend to create an account for me back in the day. Perhaps Governments need to consider legislating that social media accounts can only be created using registered emails. It doesn't stop people using different emails for communications or logging into other types of platforms, nor will it give Governments any access to social media account details, but if someone is suspended or banned, it can stop them from creating other accounts in other names. Similarly it can slow down the use of bots and the spread of misinformation.


Nice insight there CB27. Thanks for this.

Some form of AI needs to be utilised. Remember we need to prevent the hateful posts from going online in the first place that could cause harm to individuals, groups and organisations.

An anonymous account can be blocked, that does not stop that person from creating another with a fake email account. In order to stop this you need to find out the devices used to create them by Internet Protocol (IP) tracing, and network used to attach the device to the Internet (WIFI or 4G or 5G these days). The Cyber crimimal needs to be found and devices confiscated.


crissdee wrote:
I totally get your point PWUM, but I don't think it is as simple as that. To take your own example;

Prof Wind Up Merchant wrote:
For example I should not be able to post the following about anyone here.

"Person X is a queer faggot and is a danger to society"

In this example an algorithm needs to running in the background to analyse the post and identify the words "Queer" and "Faggot". If I am about to post it, a warning should pop up saying that this message cannot be posted on this website as it contravenes laws on preventing the spread of online hate.


Suppose I wanted to describe a rather unusual meal in a restaurant, and said;

"A queer thing happened last night, I ordered a serving of faggots, and when it arrived, there was only a single faggot on my plate."

Or, suppose I phrased your example like this;

"Person X is a qu**r fa**ot, and is a danger to society"

One of those posts (the harmless one) would be blocked, the offensive one would get through. The trouble with censoring language is that context is all. At the moment at least, only the human mind can identify context (unless there is AI software of which I am unaware) It is obviously impractical to have enough people employed to keep up with every strand of social media, so it must be automated. If there is such nuanced software, then of course it should be used, but I am less than convinced that it is available.


For crissdee maybe there needs to be AI to examine both the examples you gave which will identify the one about the restaurant as the non-offensive comment, and the one with the *s as the offensive one. Person using the *s can't cheat the system. The AI needs to examine the whole post then.

 
crissdee
1379419.  Wed Apr 14, 2021 7:45 am Reply with quote

Exactly my point Prof, but I am not sure that AI has reached that stage of development yet. If anyone knows better, I am all ears.

Apart from the bits that aren't ears of course....

 
technologywell
1379441.  Wed Apr 14, 2021 2:40 pm Reply with quote

Privacy is the most imp thing. You guys shouldn't share your personal images on the internet.

regards
<spammy link redacted>

 
dr.bob
1379495.  Thu Apr 15, 2021 5:25 am Reply with quote

CB27 wrote:
In each case there will then be moderators, and while social media companies don't usually employ these direct (for various reasons, but mostly to do with tax because of business status, etc), they do actually employ thousands of moderators through contractors.


One thing that often seems to be missed in discussions about policing internet content: you're right to point out that any automated system cannot be relied upon exclusively and eventually human moderators will be needed to check content.

I recently saw a TV listing for a documentary that interviewed some people who worked as these moderators. I didn't watch the documentary because just the description in the listings was horrific enough, but it essentially described the effect on these people's mental health as a result of doing a job that required them to look at pictures of extreme porn and child abuse all day every day.

While such work needs to be done, a lot of money and effort needs to be put into making sure the people doing that job are OK. Currently they seem to be being largely ignored by all parties who simply agree that "something must be done" but are not keen to get into the nitty gritty of the details about how it is done.

CB27 wrote:
It doesn't stop people using different emails for communications or logging into other types of platforms, nor will it give Governments any access to social media account details, but if someone is suspended or banned, it can stop them from creating other accounts in other names. Similarly it can slow down the use of bots and the spread of misinformation.


The major problem with internet moderation at the moment is that different parts of the world have very different problems. In the industrialised West, people are largely free to pursue a peaceful, law-abiding life and their major problem is being affected by offensive content that should be banned. Giving authorities the power to restrict people's access to certain platforms is the perfect way to tackle this issue.

Meanwhile, in places like Myanmar or North Korea, having free, anonymous access to social media is an amazingly powerful tool to help ordinary people fight against repressive regimes. Giving authorities the power to restrict people's access to certain platforms in those areas would be a catastrophic thing to do.

It's not clear how best to square this circle.

 
cornixt
1379517.  Thu Apr 15, 2021 9:23 am Reply with quote

technologywell wrote:
Privacy is the most imp thing. You guys shouldn't share your personal images on the internet.

regards
<spammy link redacted>

Funny reminder that spam is one of the main results of unmoderated forums.

 
PDR
1379522.  Thu Apr 15, 2021 10:23 am Reply with quote

Prof Wind Up Merchant wrote:

An anonymous account can be blocked, that does not stop that person from creating another with a fake email account. In order to stop this you need to find out the devices used to create them by Internet Protocol (IP) tracing, and network used to attach the device to the Internet (WIFI or 4G or 5G these days). The Cyber crimimal needs to be found and devices confiscated.


Sadly that won't do much for you. People can use public wifi (which would have its own IP) but even if they don't the IP can be spoofed, and many people use IP proxies as a matter of routine for legitimate and less legitimate purposes. We even have members of this august and law-abiding forum who use I proxies to access things like BBC iPlayer from outside the UK. If DrBob (I think DrBob is the forum IT custodian, if not then whoever is) so wished he might find that my posts to here could come from any of at least five or six IP addresses - I have two internet feeds to the household plus my phone (to which I sometimes tether the laptop when other family members are hogging the bandwidth) and until last march I spent a lot of time in hotels connected to their wifi. So that doesn't block much.

MAC address blocking might help, but these can also be spoofed/chaffed and they only identify a specific device rather than the person who uses it. The miscreant then just needs to change to another device, use a public device (internet-cafe/library/school/job-centre etc) and service is restored.

Action against the miscreant's service provider may not be practicable if they are not in a friendly jurisdiction. Good luck trying to get punitive action against a provider based in Russia, India or Nigeria.

I'm not sure what the answer is (or even if there IS one), but those aren't it!

PDR

 
crissdee
1379530.  Thu Apr 15, 2021 12:40 pm Reply with quote

Yes, I fear that we may just have to accept that, if we have a platform where we can speak freely, then everyone gets to speak freely.

 
Jenny
1379559.  Thu Apr 15, 2021 4:36 pm Reply with quote

PDR - you have in fact posted from 540 different iP addresses, 185 of which only have one post, and another 207 of which have between ten and two posts. The main address you post from records 6408 posts and the next most frequent address records 1639 posts.

 
PDR
1379563.  Thu Apr 15, 2021 5:02 pm Reply with quote

I could make a confident stab at identifying those top two, mainly because they would be the only ones with enduring static IPs. I decided to keep the previous post simple (for once) and so I didn't include the other issue - that most home-user IP addresses are not constant, but are "leased" for between one and ninety days, after which they get renewed and replaced. For most people this has no effect unless they are trying to host a website (which needs an enduring static IP so that the Domain Name Servers can index it). You can usually negotiate a static I from your ISP, but it comes with extra costs and you need to be a bit more conscientious with your firewall settings once you have it.

The other IPs will probably be mostly ones that have been my home IPs (of which there are two at any one time) and they will change every 2-3 months which is why there are so many of them...

PDR

 
kirito
1384927.  Sat Jul 10, 2021 11:54 am Reply with quote

Alfred E Neuman wrote:
My problem is that even if Iíve got one contact who stays on WhatsApp, then Iím stuck with keeping it installed. And if thatís the case, I might as well just keep using it as I do now.Visit here

absolutely...not just wapp, if one of my friend is using any other app, i will install that too to keep in touch with him.

 

Page 3 of 4
Goto page Previous  1, 2, 3, 4  Next

All times are GMT - 5 Hours


Display posts from previous:   

Search Search Forums

Powered by phpBB © 2001, 2002 phpBB Group