Bots Have Taken Over Nearly Half the Internet, but Almost a 3rd of Users Can't Tell Difference Options

[Christian J]

https://www.theepochtimes.com/tech/bots-hav...ference-5448718

[Darin McGrew]

On the other hand, surveys that intentionly ignore these responses are just creating biased irrevelant surveys.

I think it depends on what they're trying to figure out. If they want to know more about people who play video games, then it doesn't make sense to continue once you say that you never play video games. But if they're trying to figure out general attitudes towards smart home technology, then yeah, they should include people who don't use it, or who use it just a little. In that case, limiting their study to people who already have lots of smart home devices would be short sighted.

[pandy]

I've used a lot of survey sites and I know only one that feels professional and actually pay a decent hourly wage (specified). It's clearly stated who the client is, most often a university institution, you can contact the researcher and so on. The surveys can also be more taxing, but aren't always. But after spending an hour on a psych test where you stare at fast switching images with colored dots, you actually need a break!

But otherwise, I'm mostly surprised that companies actually pay for surveys that are so bad as they often are.

[pandy]

Read an article yesterday saying that AI generated child pornography increases rapidly. On top of the obvious problems with that they often use images of children from already existing real porn as a starting point. So already abused children get virtually abused again. And the police and other organizations that fight child porn need to sift through it all and try to decide what's real and what's not, which they said isn't easy to judge, so it slows them down significantly.

[pandy]

Only for Christian - unless you understand Swedish. Sorry for that.
https://sverigesradio.se/avsnitt/max-tegmar...marpratare-2023

Sommar (Summer) is a long running radio show that lets known and not so known people talk about what they want for an hour and a half.

Max Tegmark is a renown scientist with an interest for AI. I haven't listened to the end of the show yet, but it seems he thinks AI will be the end of us. Maybe a grain of salt is needed, but interesting anyway.
https://en.wikipedia.org/wiki/Max_Tegmark

[Christian J]

an hour and a half.

[pandy]

an hour and a half.

Short attention span?

Don't worry. There's music too.

[pandy]

He told an anecdote that I found interesting.

An AI asked humans for help to read a CAPTCHA. I didn't expect it would. I guess at a forum or a group chat of some kind. The person it spoke to was on to it and asked if it was bot. Then the AI lied and said it was a visually impaired person. I didn't expect that either.

But Tegmark didn't say what AI. So it could be something experimental that was programmed to do exactly that. Anyhow, it kind of shows the potential. I think Tegmark and the numerous other scientists that raise a warning flag now are right. In the wrong hands AI could be lethal.

But I don't see how we can stop it from ending up in the wrong hands. Treaties between countries are of little use when it comes to traditional war, so it hardly will work in this case. And there are always private enterprise and omnipotent individuals.

[Christian J]

He told an anecdote that I found interesting.

An AI asked humans for help to read a CAPTCHA. I didn't expect it would. I guess at a forum or a group chat of some kind. The person it spoke to was on to it and asked if it was bot. Then the AI lied and said it was a visually impaired person. I didn't expect that either.

But Tegmark didn't say what AI. So it could be something experimental that was programmed to do exactly that. Anyhow, it kind of shows the potential. I think Tegmark and the numerous other scientists that raise a warning flag now are right. In the wrong hands AI could be lethal.

Doesn't sound too bad, unless/until an AI becomes better at social engineering than even human online scammers.

I don't think AI:s need help with CAPTCHAs anymore though: https://forums.htmlhelp.com/index.php?showtopic=61360

But I don't see how we can stop it from ending up in the wrong hands. Treaties between countries are of little use when it comes to traditional war, so it hardly will work in this case. And there are always private enterprise and omnipotent individuals.

I suspect the wrong hands are the ones already creating it... Question is how much more damage Big Tech companies like Google/Twitter/Facebook can do? There are already human shills and astroturfers trying to influence forums, Wikipedia etc. Much more of the same and maybe real humans will just become more careful online?

Not sure what governments may use it for, except trying to sway public opinion (which they already do). Any ideas?

[coothead]

. Not sure what governments may use it for, except trying to
. sway public opinion (which they already do). Any ideas?

Here is just one article...

How artificial intelligence systems could threaten democracy

...written four years ago.

coothead

[pandy]

I think it could be used for a lot of things. For instance what if AI instead of generals plans a war and take the strategic decisions? Add to that cadres of drones and satellites that feed it with information. All technologies that the other part may not have access too.

I think the point of the example he used was that an AI reached out into our world and that it lied about what it was.

If we think telemarketers and phone scams are a PITA now, imagine when AI calls every person over 80 and convince them to buy this great cell phone deal or asks for their bank login and bank ID.

[Christian J]

I think it could be used for a lot of things. For instance what if AI instead of generals plans a war and take the strategic decisions? Add to that cadres of drones and satellites that feed it with information. All technologies that the other part may not have access too.

The big fear is of course allowing an AI to start wars by itself, or escalate it to say using nuclear weapons.

I think the point of the example he used was that an AI reached out into our world and that it lied about what it was.

So that AI would spontaneously make a forum post to get help solving a CAPTCHA (in turn because of some unknown spontaneous purpose)? Considering the delay before you get a reply on forums, that suggests the AI was very anxious to get past that CAPTCHA. But it's an interesting thought: a human asks an AI a question, and the AI then passes on the question to web forums (where other AI:s may reply).

How did they determine it was really an AI?

If we think telemarketers and phone scams are a PITA now, imagine when AI calls every person over 80 and convince them to buy this great cell phone deal or asks for their bank login and bank ID.

I have no problem with unsolicited phone calls (so far ). If this becomes a problem people must simply stop publishing their numbers. Once the consumer outrage becomes big enough maybe even the mobile providers may become more restrictive.

[pandy]

I think the point of the example he used was that an AI reached out into our world and that it lied about what it was.

So that AI would spontaneously make a forum post to get help solving a CAPTCHA (in turn because of some unknown spontaneous purpose)?

The story doesn't tell. But it doesn't matter really if it did it on its own or was programmed to act that way. At least I haven't thought of AI as being that proactive. But that of course, I realize now, depends on what the people behind it let it do.

Considering the delay before you get a reply on forums, that suggests the AI was very anxious to get past that CAPTCHA. But it's an interesting thought: a human asks an AI a question, and the AI then passes on the question to web forums (where other AI:s may reply).

Could have been a chat. Story doesn't tell that either.

If we think telemarketers and phone scams are a PITA now, imagine when AI calls every person over 80 and convince them to buy this great cell phone deal or asks for their bank login and bank ID.

I have no problem with unsolicited phone calls (so far ). If this becomes a problem people must simply stop publishing their numbers. Once the consumer outrage becomes big enough maybe even the mobile providers may become more restrictive.

It is a problem already. Mainly elderly people are scammed all the time, tricked to use their Bank ID during the call and whoops - there went their savings. If 10 people can scam x elderly persons in a week how many do you think AI can scam?

Phone numbers are bought and sold. The UL (local bus company) app was hacked the other week. White hat obviously, as he informed UL about the hole. He got hold of the identity of over half a million people, including email, phone numbers and buying history. No CC numbers though. Sort those people after the national identification numbers and feed the AI the identity and numbers of those born before a certain year and start the scam calls.

[pandy]

I actually have a friend whose boyfriend fell for one of those "Hello, I'm calling from Microsoft" calls. He's not old, but a pretty nervous type and no so good at English. So I imagine he was sweating from trying to express himself in English already and didn't have so much brain capacity left. He gave them his CC number and lost 37 000. He got it back more than half a year later without a word of explanation, neither from the bank nor the police.

I've got one of those calls too. It's amazing anyone can fall for it, but it happens. These new scams are much more smart than that. They don't speak broken English, they speak Swedish and convince the victim they are calling from their bank.

[pandy]

I just learnt (SVT) that on services like TikTok and Instagram every keystroke you make is recorded. You don't even need to hit the submit button. Those two services were just examples. I guess there are a lot more.

[coothead]

. I just learnt (SVT) that on services like TikTok and Instagram every
. keystroke you make is recorded.

Luckily for me, I have absolutely no connections to Social Media.

If I did then I would Copy&Paste this...

F*ck You TikTok
F*ck You Instagram

...at regular intervals.

I would hazard a guess that all appliances that are labelled as
SMART will also behave in a similar manner.

Luckily for me, I have absolutely nothing with that abominable
label.

coothead

[pandy]

Do sites like that even work of you block everything?

Luckily for me, I have absolutely no connections to Social Media.

You are here, aren't you? Per definition forums are also social media. But I get what you mean.