> What he didn't know at the time is there is no phone number for Facebook customer support.<p>Part of the problem here is that Facebook (though in fairness, they are not unique here) has left this traditional path of escalation void, leaving only fake numbers. They don't even have a real number to play a recorded message affirming that there is no ability to call.<p>ETA: For instance, I notice Facebook appears to own the typo squat `facrbook.com`. I feel like it's the same principle, though I assume toll free numbers are more expensive.
This one is pretty bad. This guy found a fake Facebook customer support phone number in a Google search, then asked the Meta AI chat in Facebook Messenger if the number he found was a real Facebook help line... and Meta AI said that it was. There's a screenshot of the chat in the article.
Seems like this information came from Quora: <a href="https://www.quora.com/Is-1-844-457-1420-really-Facebook-support" rel="nofollow">https://www.quora.com/Is-1-844-457-1420-really-Facebook-supp...</a>. Screenshot: <a href="https://postimg.cc/gallery/2nFq5Cm" rel="nofollow">https://postimg.cc/gallery/2nFq5Cm</a>.<p>I suspect the helpful SEO guy who posted this answer was trying to get more visibility on Quora so answered many questions automatically or semi-automatically without verifying anything.<p>This is the beginning of the post:<p><pre><code> Ruhul Alom
Social Media Marketer at Social MediaAuthor has 2.9K answers and 1M answer views6mo
My dear !
Yes, 1-844-457-1420 is a valid Facebook support phone number. It is a toll-free number that is available 24/7. You can call this number to get help with a variety of Facebook issues, such as:
Resetting your password
Logging in to your account
Recovering a hacked account
[...]</code></pre>
Again and again we see that LLMs are great for creative output and terrible for anything where correctness matters. You should only use it for the latter scenarios when generating answers is slow/hard/expensive, but verification of answers is quick/easy/cheap. Probabilistic and non-deterministic answers have their place, but these companies marketing them in products need to do a better job expressing the limitations.
An older client got scammed by a fake Amazon-Hotline. They bought a XBox-gift-card while on his PC via Teamviewer, till he pulled the power cored.<p>He then called me and I tried to find the official Amazon-Hotline on amazon.de. Since I was unable to find it I had to asked a search engine. The only results where third-party sites. It where from journalistic magazines I recognize (like chip.de) but still yet another gamble.
When I worked on a customer facing chatbot at my previous employer, we specifically wrote in the prompt "our customer service is not reachable by phone", and we tested that the chatbot was able to use that information and respond appropriately.<p>But I guess you can't expect a tiny startup like Facebook to invest money into having 1 employee part-time tweaking the prompt of the chatbot to respond appropriately to commonly recurring user questions.
Yes, AI in its current form is going to be a problem. I'm sure we haven't heard the worst yet. An AI may eventually kill a user.<p>I believe the heart of the problem is that corporations are riding a hype wave as long as they can, and an AI chat looks like super convincing, next level stuff thanks to the simple interface that hides the fact that you cannot communicate with this one as you would with a human being. You use natural language and it responds with natural language, which makes it not only convenient, but also dangerous.<p>There's money to gain on all this. While at the same time, hallucinations are an unsolved problem as well as making AI humble enough to realize and tell users that they just don't know. The combination of hallucinating, raising convincing arguments, being confidently incorrect, and not knowing the boundaries of your knowledge base, is a terrible one to let loose as officially sanctioned products.
One of the things about LLM-based AI that concerns me the most is realizing that the average person doesn’t understand that they hallucinate (or even what hallucination is).<p>I was listening to a debate on a podcast a while ago and one of the debaters kept saying, “Well, according to ChatGPT, […]”—it was incredibly difficult listening to her repeatedly use ChatGPT as her source. It was obvious she genuinely believed ChatGPT was reliable, and frankly, I don’t blame her, because when LLM’s hallucinate, they do so confidently.
My 89 yr old data called "AMEX" and was scammed. He googled the number for AMEX and took the top result (he says, I did not witness this). I'm across the country, so that zoom session was quite tedious (it took us an hour to get the permissions straightened out for zoom to be able to share his screen).
I'm terrified of this happening to my elderly parents. It's why, even though it can be time consuming, I always have them run "tech support" issues (no matter how small) through me or my bro in law so some foreign scammer doesn't drain their accounts.
This is the real danger of AI, forget the “singularity” or any of that sci-fi crap. AI is going to destroy the average human’s already suffering reasoning ability.
The tolerance of society for social experiments, entrepreneurial and ai is something we consider allmende, but we are currently building up a solid "anti" sentiment against all of it, liberalism, disruptive technology and i can imagine a "Luddite" party like MAGA shutting it all down hard and fast in the future. I can already imagine some future bureaucracy, evaluating any business idea suggested for scam and harm potential and ending most of them before they even start. And this stuff right here is, where it was born. The prison holding your future self, it was planted right here.<p>_____________________________________________________<p>Everything ever worth reading was written in the Pre-Collapse internet. So why not become a software-archeologist - digging for the golden past? Exhume it, get it back running, bring it all back, perfectly fine, software, books, games, our decadent ancestors abandoned and threw away to write off as rust. You too can help, rediscovering a past that worked better, untainted by AI, not yet riddled with Add-HD-Adds, when developers still had to be competent and companies still competed. Meet hot dig-site-teams near you- now. Join Past-Querries-Quary Inc. Can we dig it, yes we can!
This reminds me of the time I reported a fake PayPal email saying my account had been suspended to PayPal. The woman who answered the phone for PayPal told me very emphatically that I HAD BETTER HURRY UP AND DO EVERYTHING THEY TOLD ME TO!
I'm not a native English speaker, so I don't know how it is in any English-speaking country, but when I ask my Polish friends about the word "epistemology", they just don't know it.<p><a href="https://en.wikipedia.org/wiki/Epistemology" rel="nofollow">https://en.wikipedia.org/wiki/Epistemology</a><p>According to Google: the theory of knowledge, especially with regard to its methods, validity, and scope, and the distinction between justified belief and opinion.<p>Even though they wouldn't know the term, we all learn how to figure out what's true and what's not: we learn it when watching cartoons about lying, or when interpreting texts in school and so on. But imagine you go to a doctor, and have a small talk in which you say "I was always fascinated by medicine", to which the doc responds "What is medicine?" - you probably would run away from that doctor.<p>And yet here we are, living in the "Information Era", and yet we're still missing the very basic techniques of figuring out the truth: if you look at the statistics of religion/atheism, no group holds over 50% of population - meaning THE MAJORITY IS WRONG - and not on a nuanced thing like the majority not being able to tell the average distance between the Earth and the Moon with 1 meter accuracy. No, on something as important and world-view defining as the existence and character of God, most of us are wrong.<p>The percentage of flat-earthers in America is a 2-digit number...<p>So the problem here isn't that Facebook doesn't have a support number. The problem is much deeper, and in a way, it's good that people suffer from their stupidity: it's like programmers suffering from errors - in the end of the day they end up with their logical thinking improved. Question is: how do we reshape the society to replace production errors with compilation errors, or how do we educate ourselves to minimize the frustrating error messages.
As a millenial, I'm more amazed that someone willingly uses a phone for non-mandatory and not-burningly-urgent phonecalls... why on earth would anyone do that is way beyond me.