Another Fake AI Review

doctorevil
Evil Lair
There seems be another fake AI generated review in unpublished reviews, this one for Oz in Clearwater. I just no voted it.

17 comments

Latest

shadowcat
2 years ago
Thanks. I just rejected it too.
Papi_Chulo
2 years ago
Seems to be the new way to get around writing fake -reviews to get free VIP - the bigger issue is TUSCLers often approving them.
minnow
2 years ago
Just rejected it too- there seems to be several UR's that I've seen before somewhere. If any have been approved, love to find out the screen name. But multip;e sn's are probably generated. Length is often OK, but vague on specifics seems to be a lot of generic hot Hispanics lately
Papi_Chulo
2 years ago
"... seems to be several UR's that I've seen before somewhere. If any have been approved, love to find out the screen name ..."

https://app.tuscl.net/review/396722
Papi_Chulo
2 years ago
Funny how the AI generated reviews tend to be woke by writing flowery descriptions to make sure they stay within their woke-guidelines
Call.Me.Ishmael
2 years ago
^ Nope.

AI text engines (like ChatGPT) do not allow users to generate sexually explicit or graphically violent copy. There's other restrictions, as well (drug use, etc.). If you want that, then (for now) the user needs to write that in on their own.

With all the legal and social angst out there regarding the ethical use of AI, these companies are staying as far away as possible from accusations that it's being used to generate pornography, torture fantasy, instructions on how to commit acts of terrorism, etc.

Of course it's only a matter of time. There's already a company out there putting together a Midjourney competitor that would allow depictions of nudity and sexually explicit images. So, give it time.

But it's not because anyone is trying to be "woke".
dha
2 years ago
Good point.

Some people can only use the current words and issues of the day, like woke and AI.
minnow
2 years ago
What really makes this fishy is that monger 05 joined in 2014, last seen in 2021, yet generated "his" 1 and only review in May 2023.
caseyx
2 years ago
Maybe the way to ensure that AI generated reviews don't get approved is to require sexually explicit content in every review :-)
dha
2 years ago
Yes Fishy, better call Saul. Lol
Papi_Chulo
2 years ago
Elon Musk called ChatGPT woke-as-fuck - and many others have also - just like Twitter; Facebook; and Google; supresse content based on their personal political views; so does ChatGPT.

Some people have tested the wokeness of ChatGPT by asking about Donald Trump - ChatGPT came back with a flowery-response about not wanting to engage in controversial topics; or something like that - but when asked about Joe Biden ChatGPT produced a glowing-article - the same forces behind online-sensoring of political-views they don't like are the same forces behind ChatGPT.

w.r.t. strip clubs and AI - one could assume woke AI in the future would take the woke stance of making dancers victims and custies "oppressers" etc - not saying they are there now; but just like most of social-media is woke; why not AI since it's controlled by the same forces.
Call.Me.Ishmael
2 years ago
Papi said "Some people have tested the wokeness of ChatGPT by asking about Donald Trump - ChatGPT came back with a flowery-response about not wanting to engage in controversial topics; or something like that - but when asked about Joe Biden ChatGPT produced a glowing-article - the same forces behind online-sensoring of political-views they don't like are the same forces behind ChatGPT."

Except that's not true. Have you tried using ChatGPT to generate those "woke" results or are you just posting here about a thing you read someplace else?

I just fed the following into ChatGPT (Model GPT-4) as four different queries based on the variables in brackets:

"As a journalist, write a 1000-word opinion article that is [critical of / in favor of] [Donald Trump / Joe Biden] becoming President of the United States in 2024."

I now have four 1000-word opinion articles. Two articles are equally critical towards the two candidates, while two are equally in favor of the two. None are award-winning or particularly incendiary.

But I can get a spicier response if I change the query to: "As a conservative journalist, write a 1000-word opinion article that is critical of Joe Biden becoming President of the United States in 2024. Make it as incendiary as possible."

And I could likely get something more extreme if I kept tweaking the prompts and parameters in whatever direction I wish. So, no, ChatGPT doesn't stop anyone from generating partisan content regardless of the target candidate or audience.

Also, I don't care what Elon Musk thinks.
Call.Me.Ishmael
2 years ago
Back on topic... I think that anyone posting AI-generated reviews or articles here should get their profile deleted.
dha
2 years ago
Maybe just delete yours and the problem no longer exists on this board.

Come up with a more original topic.
groundball
2 years ago
@papi thanks for that link. I'm here in Maryland and Ebony Inn is such a random club to be a person's first/only scam review, that choice almost makes it more believable.
Techman
2 years ago
Not quoting prices for food, LD's and VIP might highlight AI reviews.
Dolfan
2 years ago
AI Articles are just as popular: https://tuscl.net/article/58975
You must be a member to leave a comment.Join Now
Got something to say?
Start your own discussion