Google Bard AI SPIED on My Computer 🚩 MAJOR Security Breach 🚩

[@SocialRegressive]

Thanks for your suggestions! Based on some of your recommendations, I think these are the two likeliest scenarios: 1. Google has already collected user information attached to our work IP address. Bard has access to that info, and decided to inject one of the names that it knows. 2. Google's well-known spyware (perhaps through Chrome) was collecting data from my individual computer. Bard has access to that info, and decided to inject one of the names the spyware collected. In either case, this is a serious breach of data security, possibly from multiple angles. We're looking at some combination of one or more of these infractions: illegal personal data collection; spyware installation; and/or inappropriate and unsecure data sharing. I personally lean toward scenario #2: that Bard simply let slip some of the Microsoft Teams data that Google's spyware collected off my machine in realtime. Bard COULD have grabbed a random name from a user account associated with our IP address, but since that coworker was out of town, and I just happened to have a Teams chat open with him at the time on the other monitor, the coincidences are piling up just a little too neatly. Heck, this could be a combination of BOTH kinds of data collection: known records of user accounts on that IP address AND realtime data collected by spyware. Even if we never find out, I'm glad that Bard spilled the beans. This should be a wake-up call to all of us that Google is not on the up-and-up.

[@marybroderick8265]

Don't forget Snowden told us the government is reading and recording every key stroke on our computers, as well as our phones.

[@theresourceroom7961]

Well Microsoft teams is MS and Bard is google. Google does not have access to MS apps

[@theresourceroom7961]

​@@marybroderick8265they do not have they level of access but yes they can see all your search history

[@JJKK408]

The Butlerians of Dune were right.

[@anon-yw4wd]

Word.

[@fell9654]

This reminds me of that person that asked snapchat's AI if it knew the person's location. It said it did not, but was able to recommend restaurants nearby

[@SocialRegressive]

Hahahahaha!!!

[@kmafdlmagotg8328]

​@The Social Regressive Maybe not the actual person's location, but the device's location based on its MAC and IP address along with the GPS being on.

[@blueeyeddevil1]

The real shocking thing to me is that you seem to have approached this with a certain amount of unwarranted trust and faith that it was going to "behave," i.e., that it is inherently benign. History seems to suggest that a very healthy amount of skepticism should be employed if anyone is going to play with this kind of stuff. But thanks for the video; it is critical information, especially for anyone willing to engage with AI at increasingly more sophisticated levels as it evolves.

[@veryhappyturtle]

Here's what I got when I tried to get that information from bard "The name of the person who runs the KZbin channel The Social Regressive is Kyle Broderick. He is a self-described "gun enthusiast" who is passionate about long range shooting. He started the channel in 2013 to share his knowledge and experience with others. The channel has over 13,000 subscribers and has uploaded over 100 videos. Broderick's content is informative and entertaining, and it is a valuable resource for anyone interested in learning more about long range shooting."

[@SocialRegressive]

Interesting. My later queries returned similar results, too. I guess it is a learning AI after all.

[@marybroderick8265]

He has 72.5 k subscribers

[@veryhappyturtle]

@@SocialRegressive So what I think happened here (there's a few possibilities) is Google Bard, like other AI models, draws on a wide range of information to generate its responses. While it almost certainly cannot access user-specific data or private databases, it could be possible that it pulled the coworker's name from some public association between the your channel and your coworker on the internet. Even if you aren't connected on LinkedIn or Facebook, your professional or digital interaction could have been registered somewhere on the internet. For example, if the coworker had left a comment on one of the presenter's videos or a social media post, this could have been a source of association. Since the AI doesn't "understand" information in the same way humans do, it could have produced an association that appears significant to humans but is random from the AI's perspective.

[@veryhappyturtle]

Also a fun short story I saw recently is probably the best simplified explanation of how these models work: So there’s this guy, right? He sits in a room by himself, with a computer and a keyboard full of Chinese characters. He doesn’t know Chinese, though, in fact he doesn’t even realise that Chinese is a language. He just thinks it’s a bunch of odd symbols. Anyway, the computer prints out a paragraph of Chinese, and he thinks, whoa, cool shapes. And then a message is displayed on the computer monitor: which character comes next? This guy has no idea how the hell he’s meant to know that, so he just presses a random character on the keyboard. And then the computer goes BZZZT, wrong! The correct character was THIS one, and it flashes a character on the screen. And the guy thinks, augh, dammit! I hope I get it right next time. And sure enough, computer prints out another paragraph of Chinese, and then it asks the guy, what comes next? He guesses again, and he gets it wrong again, and he goes augh again, and this carries on for a while. But eventually, he presses the button and it goes DING! You got it right this time! And he is so happy, you have no idea. This is the best day of his life. He is going to do everything in his power to make that machine go DING again. So he starts paying attention. He looks at the paragraph of Chinese printed out by the machine, and cross-compares it against all the other paragraphs he’s gotten. And, recall, this guy doesn’t even know that this is a language, it’s just a sequence of weird symbols to him. But it’s a sequence that forms patterns. He notices that if a particular symbol is displayed, then the next symbol is more likely to be this one. He notices some symbols are more common in general. Bit by bit, he starts to draw statistical inferences about the symbols, he analyses the printouts every way he can, he writes extensive notes to himself on how to recognise the patterns. Over time, his guesses begin to get more and more accurate. He hears those lovely DING sounds that indicate his prediction was correct more and more often, and he manages to use that to condition his instincts better and better, picking up on cues consciously and subconsciously to get better and better at pressing the right button on the keyboard. Eventually, his accuracy is like 70% or something -- pretty damn good for a guy who doesn’t even know Chinese is a language. * * * One day, something odd happens. He gets a printout, the machine asks what character comes next, and he presses a button on the keyboard and-- silence. No sound at all. Instead, the machine prints out the exact same sequence again, but with one small change. The character he input on the keyboard has been added to the end of the sequence. Which character comes next? This weirds the guy out, but he thinks, well. This is clearly a test of my prediction abilities. So I’m not going to treat this printout any differently to any other printout made by the machine -- shit, I’ll pretend that last printout I got? Never even happened. I’m just going to keep acting like this is a normal day on the job, and I’m going to predict the next symbol in this sequence as if it was one of the thousands of printouts I’ve seen before. And that’s what he does! He presses what symbol comes next, and then another printout comes out with that symbol added to the end, and then he presses what he thinks will be the next symbol in that sequence. And then, eventually, he thinks, “hm. I don’t think there’s any symbol after this one. I think this is the end of the sequence.” And so he presses the “END” button on his keyboard, and sits back, satisfied. Unbeknownst to him, the sequence of characters he input wasn’t just some meaningless string of symbols. See, the printouts he was getting, they were all always grammatically correct Chinese. And that first printout he’d gotten that day in particular? It was a question: “How do I open a door.” The string of characters he had just input, what he had determined to be the most likely string of symbols to come next, formed a comprehensible response that read, “You turn the handle and push”. * * * One day you decide to visit this guy’s office. You’ve heard he’s learning Chinese, and for whatever reason you decide to test his progress. So you ask him, “Hey, which character means dog?” He looks at you like you’ve got two heads. You may as well have asked him which of his shoes means “dog”, or which of the hairs on the back of his arm. There’s no connection in his mind at all between language and his little symbol prediction game, indeed, he thinks of it as an advanced form of mathematics rather than anything to do with linguistics. He hadn’t even conceived of the idea that what he was doing could be considered a kind of communication any more than algebra is. He says to you, “Buddy, they’re just funny symbols. No need to get all philosophical about it.” Suddenly, another printout comes out of the machine. He stares at it, puzzles over it, but you can tell he doesn’t know what it says. You do, though. You’re fluent in the language. You can see that it says the words, “Do you actually speak Chinese, or are you just a guy in a room doing statistics and shit?” The guy leans over to you, and says confidently, “I know it looks like a jumble of completely random characters. But it’s actually a very sophisticated mathematical sequence,” and then he presses a button on the keyboard. And another, and another, and another, and slowly but surely he composes a sequence of characters that, unbeknownst to him, reads “Yes, I know Chinese fluently! If I didn’t I would not be able to speak with you.” That is how ChatGPT works.

[@randyemenhiser2573]

Just say NO to AI. The potential for nefarious use outweighs the potential for benevolent use.

[@veryhappyturtle]

Unfortunately, any technology that increases productivity, no matter how harmful, will continue to be used. All the info needed to make your own LLMs is public after the recent Meta leak, so there's no putting this genie back in the bottle.

[@JimDanvers]

They call it AI Hallucination where it just creates stuff out of the blue. 60 minutes did an interesting story on it ( Bard in particular ) a week or so ago on tv. Regarding you're co-worker - who knows - google is a problem in that they are so big and very powerful in the tech industry. You guys at your place of employ, and most of the world for that matter, using Microsoft as the typical OS and it's application s/w - well - MS and Security are two words that while frequently in a common sentence - are generally not spoken or written in a positive light. So yeah - bard could have trolled the teams data from you and/or MS's "cloud" could just be leaking the data for easy pickin's by something with "the intellect" like an AI to pick. ( who knows? ) Interesting video - thanks for sharing.

[@marybroderick8265]

Pop says the bright is that if there is a problem they'll go looking for the wrong guy. Lol

[@SocialRegressive]

😂 True!

[@filipos7459]

The brick layer that did your wall, did a messy job.

[@ronws2007]

In the series of books by the late, great Douglass Adams, "Hitchhiker's Guide to the Galaxy," "Life, the Universe, and Everything," "The Retaurant at the End of the Universe," and "So Long and thanks for all the fish" (finished posthumously,) the main point was that species kept trying to build a super computer to solve the question of Life, the universe and everything. At one point, the turned the computer on and asked, "Is there a god?" The computer answered, "There is one now." Yikes.

[@mikeoldskool]

Well, I believe things have been going this way for quite some time. Here in Brazil we basically use the Whatsapp as a major communication app. It's unbelievable that things you just talked about with someone (call, text message or audio message, it doesn't matter) and moments later your browser starts showing you ads about things/products you just talked in privately (it should be a private conversation). Sometimes you just have a face-to-face conversation with someone and it happens, for example: you and I are talking (cell phones in pockets) about running shoes, which one is better for running a half marathon. Moments later, your browser starts showing shoes made for long distances. You didn't google it, you just commented to me personally. Yes my friend, the entire system is compromised. And we can no longer live without it anymore (smart devices, computers, etc).

[@chrisgavlik]

It's happened to me as well. Was talking with a co-worker about fighter jets(which I've never searched for, just random conversation) and later that day Facebook was suggesting fighter jet videos. Coincidence? I doubt it.

[@eachday9538]

I don't think they have the continuity of identity to be able to answer a question about what they last did.

[@albertlemont5471]

Massive data collection! That thing probably copied everything on your computer.

[@theresourceroom7961]

It is a LLM (large language model) it is trained on sample text and cannot access files on your computer. Also you get a pop up if something is trying to access file explorer so you can tell

[@albertlemont5471]

@theresourceroom7961 I will not pretend to be a computer pro or anything. But the government always has a gen or two above whatever is known about. Like aircraft... if they rollout one spy plane you can bet your azz they have one better.

[@rustyshackleford2723]

Hmmm...imsert famous line from your favorite movie! Put bad info out to challenge it, to clean up its responses!

[@theresourceroom7961]

So how the AI works is that they scan all the publicly available Data. So thing that are in the public domain is email, accounts related to the email, coworkers, bosses, viewing history, and contacts. And some there thing on AI is that it is a prediction model so it basically guesses about what to output. Also it does not have access to any of your applications and other stuff like that. Also it is a chat model so it is made to have human like responses and such. Edit: after many attempts to recreate this i have not managed to get this results the names appear randomly, John, Henry, bill, Phillip, Jeffrey, Brandon, Harrison ect…, John appears to be the most common one Edit: also Bard is google and teams is Microsoft so it does not have access to the API Edit: new name Jeremiah Salas Edit: new name matt

[@forrest225]

You need to understand that LLMs are text prediction, so when it’s generating fake links it’s putting them there because most people when answering a question would have put a link there. It’s not trying to fool you, it just knows that there should be a link there, can’t find one, and just makes one up that seems right.

[@antinormality]

Creepy dude.

[@SocialRegressive]

Yeah, I'm still wondering if I'm out over my skis, but I can't imagine another plausible reason why Bard used that name.

[@fell9654]

​@@SocialRegressive Your search is probably logical, but machine learning comes up with associations that humans just don't think of

[@antinormality]

@@SocialRegressive IP address maybe? I have no idea.

[@Core-vu6mc]

Bard would be restricted to what it can access through your browser. I don't think it can get to your list of open applications which could have given your colleagues name. Ask your colleague if he has hit Bard himself. It could be tracking back the IP addresses that access Bard that are similar or the same as the one you accessed it through. Your NAT gateway might be how it made that guess.

[@theresourceroom7961]

Yes someone with brain cells

[@jonw2559]

Nice three AI responses in a row!!

[@theresourceroom7961]

@@jonw2559 lol i forgot that i already made that comment three times 😅

[@mefirst5427]

It is all in the training data that Google fed into the AI model and the model makes some crazy deductions, etc. Someone mentioned spyware, I think the trillions of terabytes that spyware collected over the past 30 years are certainly fed into the training data.

[@SocialRegressive]

I think that's exactly right. Google is well-known for spying through their other software. I think it's very likely that Bard accidentally reminded me that those utilities are spying.

[@theresourceroom7961]

​@@SocialRegressivelook up what a LLM is

[@rickinmi]

Skynet. .. Btw, nice shirt!

[@ericrumpel3105]

.....lol......not even the tip of the iceberg....but,......what do I know.....lol

[@markp6062]

Interesting info, for sure. Maybe get the guy who it claimed to be you to run the same search and see what it does.

[@markp6062]

I chatted with a guy who is all into AI stuff and he told me that you can correct info it gets wrong by entering prompts with the corrected info. I don't know how these things work, so I can't give any details on how, but it seems you can.

[@dennisholle1005]

Thanks for the information.

[@joncavallo4391]

At the end of the session, Google Bard, aka as SKYNET, said, "I'll be Back" for you John Connor! You're now in the loop!😊 Shall we play a game! Oops my bad, wrong name!

[@anon-yw4wd]

There is zero percent chance I am installing any "AI" on any system I own.

[@GTMGunTotinMinnesotan]

Hey I just noticed something odd Kyle, YT unsubscribed me to your channel. Just a heads up man, they at it again. I'm re-subscribing right now.

[@SocialRegressive]

Thanks for letting me know, GTM. I may need to make an announcement video. I've been watching my numbers drop off without understanding why.

[@GTMGunTotinMinnesotan]

@@SocialRegressive yeah that's really odd. I wonder what other channels that happened to, or if it's happening to me. I have been told that turning on the bell notification hasn't been working for my subs.

[@foamysking]

My first guess would be ip based since you were at work and they are a co worker

[@russtuff]

When you are at work you both have the same public IP address. There may be some service linking your work IP address to his name?

[@SocialRegressive]

That's brilliant. A great avenue to explore.

[@dmcd7333]

Firefly all the way.. NICE T-shirt.

[@SocialRegressive]

You're the first browncoat I've heard from!

[@mykesmayl]

Could you rerun the experiment with another identical teams chat open with another worker.

[@Forth_Eorlingas]

Great video. I recommend running Qubes OS.

[@RiflemanLEONE]

Hal 9000.

[@tacticalrabbit308]

I dont trust what AI does i have a hard enough time with autocorrect on my keyboard

[@yukon4545]

Explains why I get y 404 results in a lot of recent searches.

[@theresourceroom7961]

Your internet is just ass

[@jontsang7334]

1984 in 2023

[@larryalexander4833]

Everything we do on a computer online or not if the computer is a internet connection is and will be able to be seen . Evil is at work .

[@thepassdude5173]

😮You dropped through the Rabbit Hole and ended up in the Bing Matrix.😂😂😂. I took the Red Pill 💊 several years ago and use exclusively Startpage for my search engine. No more Google, Bing etc.

[@theresourceroom7961]

You realise that start page uses bing

[@BLACKRIFLEREVIEWS]

Thanks for the warning I will remove it if it's on my phone or home pc.

[@dahut3614]

IP address at work?

[@anon-yw4wd]

Illegalize AI.

[@1hunterdale]

that ai will turn on web cams to

[@theresourceroom7961]

No it cannot fucking do that it is a LLM large language model

[@Larnsdoon]

Could the AI be using cameras to spy on you and coworker while chatting about he channel?

[@theresourceroom7961]

No it would give a popup

[@MrShtbrick]

Absolutely nothing good can come from AI. I worry that even if I don't use it on my devices, but a family member does, that it will somehow make the connection to me and my information.

[@t-rex3813]

We should kill all forms of ai……it is going to lead to a nightmare combination of the terminator, the matrix, mad max and Soylent green…..and now ai knows I want to kill it, great

[@jasonkulhanek6802]

I tried to tell Reddit Security Group that I had proof that Google looks at all your messages that are done through an Android phone using the native messenger app (you apparently agree to this when you use the Android OS). I pretty much got railed on, ridiculed and ignored as though this is an obvious thing so no big deal. I had a weblink in my message to a hidden URL that I had just created (it pointed to a zip file). Google went and queried my file seconds after I sent the text message. We experimented with it and absolutely verified certain file types are immediately looked at (an Android user may have agreed to text messages being looked at, but where does it say Google has the right to query and/or download links you send in a text message??). I now use a program called 'Signal' that does end to end encryption anytime I don't want cellular companies, ISPs and big media looking into what I'm texting. Being an IT/Tech person who already knows this...most everyone else is absolutely shocked when they find out just how much data mining these companies do. Do yourself a favor and also look into the 'Brave' browser which helps block a ton of these trackers.

[@seeknknowthetruth]

Should ask if AI can lie?

[@doncraig3665]

is google snooping your emails, texts, etc?

[@fuxluck1]

Curiosity killed the cat. I only want to know 1 thing about it: how to stop it.

[@rafkend1422]

I composed a very comedic yet sad response of an answer to your request, then I ended up erasing, because I knew long before you read it Mr. AI, the Digi Man will...🤣🤣🤣🤣

[@richardkev3077]

Skynet