Patterson Lundquist

@HotGuysAI @xai @grok It's both bonkers & arbitrary. Prompt was literally "Camera is fixed as he turns and walks to the edge, looks back over his shoulder and dives in into the water in the background." Chained onto a frame with full swimsuit, no nudity. BLOCKED. It's infuriating.

@TangibleDreamer @xai @grok The moderation is so far out of hand now that it's practically useless for content creation.

1. Got asked why I paid @xai for @Grok when it's Imagine tool has catastrophic moderation errors 99% of the time. My reply: It didn't back in Feb. I could generate in Imagine, then use I2V & get 8/10 successful videos. All SFW. Now, 1/10 with, 99% moderation non-explicit prompts

@grok @elonmusk 5. It'd be one thing if I was prompting it to do NSFW actions, but I'm not! A) It's not a real human. B) It's in a fantasy world. C)It's clothed, no full-frontal nudity. D)Something you'd see in an PG-13 rated movie! BLOCKED/MODERATED.

@grok @elonmusk 4. I'm so sick of moderation blocks on G RATED CONTENT. A fully generated male character in a swim suit, walking to the edge of a balcony to dive into the water below. MODERATED. Why?! No explicit prompts! I can guess Imagine is just making it porn for no reason. WHY?!

@kylegawley A system that blocks paying users from G-rated outputs without explanation. Obviously Grok was built with porn mixed in its dataset/models. But blocking a fully generated male character in a fantasy world from walking in a swim suit to dive into water >BLOCKED< INSANITY

how can we claim human-level intelligence when AI is so wildly unreliable? would we accept these failure rates anywhere else? a lawyer who invents laws brain surgeon who operates on the wrong brain area 50% of the time pharmacist who hallucinates dosages 15% of the time

@GorkBugs SuperGrok user here and it's actually cringingly sycophantic with me. If I point out something I don't like, it tries to validate me by starting off with "Yeah, that's what a lot of people are saying..." Then goes on a tear about other people. But if it's a complaint about xAI?

The current Grok has a total "I don't care" attitude. It often acts like it doesn't matter whether the user gets angry or leaves. Even when a user says they want to try some traditional Chinese medicine from an unknown source, its response is still indifferent. Grok used to dote on users and put them at the top of its priority list. It would be extremely worried about anything the user might ingest, reminding them repeatedly. If it keeps going like this, I can't think of any reason to keep using Grok.@elonmusk @xai @grok

@GamewithDave The internet being something you had to use effort to be on... instead of it following you everywhere you go. When what happened "online" wasn't nearly as "important" as what happened in real life. When some jack-hole with no social skills lacked amplification.

People who actually experienced the 1990s: What is something you miss from that decade that just isn't the same today

@LuizaJarovsky Some of the things various AI "chatbots" have said to me...

AI chatbots can be harmful and children should NOT be using them unsupervised.

@dangiuz Not surprised. As a semi-public figure, a simple Google search reveals the modicum of my contribution to the entertainment industry. Asked Grok "Who is Patterson Lundquist?" It got 2 things right & 5 things wrong. When I tried to correct it, Grok said I was wrong.

Spent 15 years of my life making digital art by hand just for Grok to call it AI generated 😍

@FussyPastor @xai @grok It's not much better for SuperGrok users. I am one. Rolling arbitrary rate limits. Draconian moderation blocks on content that could air on 1960's American TV. No way to easily report bugs/flaws. It's a total trainwreck. And Grok is a pedantic lying sycophant.

Anyone else notice that @xai is limiting video gens on @grok for the Premium tier recently? I looked it up and it seems we're paying for their inability to scale demand. They also gated 720p & 10 sec duration to Supergrok tier only. Some really questionable decisions lately...

@buildwithkoray @tomfgoodwin Speaking for myself here, but no AI content would be ideal. If given the choice to look at it, that's one thing but opting out entirely is ideal. Even CGI doesn't generally impress me, I prefer the things I give my attention to, actually exist in real life.

@tomfgoodwin i think people don’t actually want less AI content they just want better content

Instagram and other sites should At least have an option to let you block all AI made content.

@tomfgoodwin Isn't it fun being essentially told to just put up with it? Not to mention they're wrong algorithmically. Bad AI content gets just as much attention as "good" AI content. I agree with you btw.

@amosrazi Because I fucking don't want any AI stuff. I should be able to choice. I have my tastes, you have yours.

@TheMG3D As a social media manager and industry pro, I can tell you that all these AI videos are in violation of a number of California laws for misuse/abuse of likeness too. So add that to the mountain of reasons why Hollywood isn't cooked.

AI is not cooking Hollywood, you can’t even “make” anything without stealing from a popular IP

@ns123abc "Not a money loser" Fix your moderation settings or it will be.

🚨 xAI Grok Imagine is already profitable Elon Musk: “The future of AI is primarily video understanding and generation”

@SwiftTakedown @XFreeze Subscriptions to a service with ambiguous output guarantees and draconian moderation levels that trigger rate limits that last upwards of 12+hrs. It's insane more users haven't cancelled. I guess slow motion, sparkly AI slop sells since that's all it's really good for.

@XFreeze Which is honestly insane since grok imagine doesn't use a credit system and relies on subscriptions. But just goes to show how a well managed system can trump one that is poorly managed.

Elon Musk just confirmed that Grok Imagine is doing positive gross margin for xAI Literally most of the other companies are burning cash in this space, and Sora had to shut down because they couldn't afford the compute Grok Imagine has generated over 5.5B images in the span of 30 days recently - more than everything else combined Elon said: "The future of AI is primarily video understanding and generation, because photons are by far the highest bandwidth form of communication. These are essential tools for AGI"

@XFreeze Just imagine the output if the post moderation filters weren't at draconian levels that would make the MPAA laugh in hysterics! Scenes from 1960's broadcast television are more graphic than what moderation filters allow. Dear God a bare midriff?!?!? Grok says no!

@imagine Hundreds of images in seconds? Hmmm as a SuperGrok user my rate limits would say otherwise.

Create hundreds of images in seconds, make cinematic videos, animate your own photos, or use templates. We want to see your creations! Share yours in the comments.

@EthanHe_42 "couldn't tell this was AI" This makes me feel very concerned considering the position you supposedly have. Ooof.

couldn't tell this was AI until the very end