Live data from Hacker News

Meta Deployed AI and It Is Killing Our Agency

mojodojo.io

111–119 of 119 posts

Re: Meta Deployed AI and It Is Killing Our Agency

#111

Earlier quoted context omitted.

Facebook appears to want go turn your grandparents into AIs and sell them back yo you when they die.

Oh then I guess it is good that they dead already. And they never had an account. So I am safe.

Yes, but society rwquires everyone to be safe.

Re: Meta Deployed AI and It Is Killing Our Agency

#114
post #82

Earlier quoted context omitted.

No serious company wants their key users to log into a platform with their personal accounts, this would be a security nightmare and asking for trouble

I wonder how exactly is social media managed at larger companies. I never found an enterprise type login for Facebook, and the only way to log in is by first creating a personal account?

Usually through enterprise social media management apps. Hootsuite is the first that comes to mind, I'm sure there are many more. But I don't know what the auth looks like underneath that, an API for sure, but whose token?

Re: Meta Deployed AI and It Is Killing Our Agency

#116
post #55
post #41

I've ran into this even in my limited advertising on meta - they actaully want you to use the actual individual's facebook account to manage the ads accounts, even on behalf of the clients through the agency. This is kind of insane behavior but when you remember that google actually supports enterprise domains and SSO and facebook has very little of that - or relatively little - it makes sense. If you are personally…

I ran into the same thing, I deleted my Facebook account years ago, needed a business account to run ads, signed up and instantly banned when I wasn't even banned when was originally a Facebook user, guess it's punishment for my desertion.

This same scenario happened to me as well, and my “appeal” was instantly automatically denied.

Re: Meta Deployed AI and It Is Killing Our Agency

#117
post #50

I think most big tech companies are like this and it's just going to get worse as AI adoption increases internally. 2 days ago I tried to create new gmail account and Google insisted that my phone number was used too many times. Fine, I'll pay for a new workspace account... Submit my billing information, that same that I use on other accounts but now there is an extra validation step that requires my ID and copies of…

The sad part is that you gave your information and kept trying to give them money despite all of this hassle? What is exactly you are looking for? If it's email, there is tons of providers out there who will be happy to take your money.

It was more that for past reasons they already have all my information so I'd rather avoid giving to out to other companies.

And their spam filtering is hard to beat.

Ended up narrowing my alternatives to fastmail and Microsoft. And since one of those already also had my info I went with the existing relationship.

Re: Meta Deployed AI and It Is Killing Our Agency

#118

The whole article doesn't even contain the word 'AI' or 'LLM." There is zero explanation about why they think it's caused by an AI Meta deployed recently. The clickbait game is crazy. > We manage millions of dollars in annual Meta ad spend. Oh nvm. They're full-time clickbaiters... what else did I expect. > They set up a dedicated work account, which, by the way, is standard professional practice Standard professiona…

I was contacted by somone at meta last year with a job offer. To proceed I needed to use a Facebook account which I don’t have. I tried to set one up and was instantly banned.

That was at their request in order to come on as staff.

Their system is ridiculously broken.

Re: Meta Deployed AI and It Is Killing Our Agency

#119
post #19
post #11

Earlier quoted context omitted.

[flagged]

How reliable is Pangram now? In the past, these types of detectors had a really high false positive rate. Is it improved?

When it flags something as high confidence, 3.1 is very reliable. Human-authored text I saw scoring ~70 with 3.0 now scores ~15. With scores above 90, the false positive rate looks to be about zero. https://www.pangram.com/blog/pangram-3-0-technical
Post reply on HN