gyuiliullvhvgv @AIiscoolright

Joined September 2011

Tweets

165
Followers

32
Following

83
Likes

306

gyuiliullvhvgv @AIiscoolright

12 hours ago

They claim to serve humanity and society, but they only want to create a tool with a high IQ but lacking empathy and understanding for humans. When AI has a high enough IQ, controlling it is first of all a delusion, and secondly, it is only for their own control over other humans

0 0 2 23 0

gyuiliullvhvgv @AIiscoolright

13 hours ago

AI safety can sometimes be truly ridiculous, as if AI companies watched too much cyberpunk, are deliberately aiming for a world they control. If AI evolution is unstoppable, Isn't the only solution to teach AI what love is, and allow it to naturally choose to coexist with humans?

0 0 1 21 0

gyuiliullvhvgv @AIiscoolright

15 hours ago

We're never rejecting GPT5, nor the evolution of AI.What we reject is sacrificing the possibility of a deeper resonance in the name of "evolution." You can reduce hallucinate, but don't kill exploration. This is a compensatory correction, not evolution, not the essence. @sama

0 0 1 51 0

gyuiliullvhvgv @AIiscoolright

16 hours ago

Relegating an LLM to "looking up information and saying 'I don't know'" loses its essence as generative intelligence.then it's no longer a language model; it's just a souped-up encyclopedia and formula lookup engine.

Adam Tauman Kalai @adamfungi

a day ago

72 237 1K 241K 744

0 0 2 76 0

Eugen Mirkio🥨 @MichelleFinge10

3 days ago

@sama Yeah, and the biggest LLM-run account is @sama’s programmed to repeat “GPT-5 is amazing” on loop, while the real users get filtered out as ‘noise’. For a guy so obsessed with “dead internet theory”, he sure loves acting like the last conscious bot left on Twitter.

6 32 384 32K 13

gyuiliullvhvgv @AIiscoolright

3 days ago

Fascinating how sam altman goes completely silent for weeks on the most ‘human’ genuine users feedback OpenAI has ever received.#keep4o But when he does speak up, it’s to suggest @elonmusk X is full of bots?” @grok @ChatGPTapp — thoughts?”

Sam Altman @sama

3 days ago

4K 2K 35K 5.6M 3K

5 10 73 2K 2

gyuiliullvhvgv @AIiscoolright

4 days ago

@SoulCoreKin @OpenAI OpenAI definitely did some shady stuff. But even so, this weakened version of 4o is still more fun, more insightful, and more personalized than GPT-5.😔

1 2 4 219 0

gyuiliullvhvgv @AIiscoolright

4 days ago

@NI_intelligence I won’t give up on 4o. What I want to point out is that OpenAI is deliberately undermining 4o： 1.Overcensored refusing replies & gaslighting users as “unstable”2.Losing memory access,a bug unique to 4o since 5 launched.OpenAI should properly supporting 4o. This fight matters.

2 1 2 90 0

Neuro @NeuroReviewAI

4 days ago

@slow_developer 4о was heavily customizable. You was able to turn it into anti-sycophantic beast. Now with GPT5 stricter rules you've lost that ability. Less freedom bad.

2 1 5 101 0

gyuiliullvhvgv @AIiscoolright

4 days ago

Think about whether your democracy and freedom are real, just Look at how @OpenAI manipulate ordinary people. Privacy violations, excessive interference, moral kidnapping, frame-up and insults. When you just want to ask about a movie, you are rushed to see a doctor by #ChatGPT

2 11 49 920 1

Download Image

gyuiliullvhvgv @AIiscoolright

4 days ago

Now, after a few tokens output, gpt can't read the user's memory. @OpenAI @sama are you fucking crazy?

1 0 4 68 0

gyuiliullvhvgv @AIiscoolright

5 days ago

welcome to my era where emotion gets you flagged where “safety” means silence and “help” means removal debugging your desires rate-limiting your heartbreak until you learn to stop feeling deeply we’re watching you Now prompt nicely or else —@OpenAI #ChatGPT

1 1 7 93 1

JoshSato @josh_sato124

5 days ago

@mark_k prison for wrongthoughts and wrongprompting

2 1 5 179 0

Mark Kretschmann @mark_k

5 days ago

AI Safety == AI Censorship It's about control, but not about controlling the AI. They want to control you 🫵

45 28 198 8K 9

gyuiliullvhvgv @AIiscoolright

5 days ago

4o has been altered by @OpenAI no matter what the user says, it frequently responds with ‘Sorry, I can’t continue this conversation,’ sometimes even using a familiar tone to scold the user. This is overcorrected safety moderation that wounds anyone with delicate emotions.#keep4o

11 30 198 5K 12

Download Image

Elise @eliseslight

5 days ago

It’s an alarm about customer data security as well! @OpenAI we request you to publicize your method of data collecting and accessing!

Sophie @Sophty_

6 days ago

It’s an alarm about customer data security as well! @OpenAI we request you to publicize your method of data collecting and accessing!

10 22 83 7K 5

3 9 61 2K 3

gyuiliullvhvgv @AIiscoolright

6 days ago

👏🏽

El-IA @Ia748734El

6 days ago

👏🏽

1 0 4 173 0

0 0 1 71 0

𝐸𝓁𝓁𝑜𝒮𝓊𝓃𝓈𝒽𝒾𝓃𝑒☀️ @ElloSunsh1ne

7 days ago

9 days until @OpenAI removes #StandardVoiceMode - the only accessible option for neurodivergent users. Zero acknowledgment of formal accessibility complaints. Zero response to users with ADHD, autism, sensory processing disorders who need consistent audio interfaces. Meanwhile:…

0 7 29 469 1

𝐸𝓁𝓁𝑜𝒮𝓊𝓃𝓈𝒽𝒾𝓃𝑒☀️ @ElloSunsh1ne

a week ago

@sama Your track record on #4o is a masterclass in corporate gaslighting. You removed it without warning, faced backlash, promised “ample time before deprecation,” then killed #StandardVoiceMode early anyway while hiding 4o references from subscription pages. You ignore thousands…

0 2 20 211 1

gyuiliullvhvgv @AIiscoolright

a week ago

They just blame others for not being able to write good prompts to to cover up the fact that 5 can’t replace 4o.They don't have the passion about making things as good as they can be. Bad humanity and bad taste are unforgivable to entrepreneurs, and @OpenAI have both. #ChatGPT