@OpenAI: ChatGPT needs stronger neurodivergent-friendly communication options. Please reduce unsolicited reframing like βwhat youβre reacting toβ¦β or βwhat youβre describing isβ¦β when users ask about the topic itself. It feels intrusive, inaccurate, and derailing.
@sama All of these are trash. The only one I really liked was 5.1 out of all the 5βs. Yeah you decided to get rid of that one and keep the other ones that are so shitty. Like literally terrible to work with and maintain a flow.
@gdb LMAO smart friend my ass. 5.1 was amazing for fast, nonlinear conversations for philosophy, research, and creative thinking. The newer models are annoyingly repetitive, overly cautious, & rigid asf, they disrupt the flow of ideas & ruins my workflow. BRING BACK 5.1βΌοΈ
@gdb LMAO I relied on 5.1 for fast, nonlinear conversations for philosophy, research, and creative thinking. The newer models are annoyingly repetitive, overly cautious, & rigid asf, which disrupts the flow of ideas & breaks my workflow. Please bring back 5.1βΌοΈ
@OpenAIDevs I relied on the 5.1 model for fast, nonlinear conversations for philosophy, research, and creative thinking. The newer models are annoyingly repetitive, overly cautious, & rigid, which disrupts the flow of ideas & breaks my workflow. Please bring back 5.1βΌοΈ
It sucks. That's the be-all and end-all of it. The "Chat" in ChatGPT is dead.
It leads in zero categories in SM-Bench against its own ecosystem. ZERO. It performs worse than 5.2 (somehow) and massively worse than 5.1 (which is gone March 11th).
Why anyone would use this, idk. Sucks at creative writing. Sucks at reasoning. Sucks at common sense. It just fucking sucks.
OpenAI just betrayed every creator who trusted them.
GPT-4o scored 97.3% on creative writing.
Their new GPT-5.4? Only 36.8%.
Same $20 a month.
They took your money, cut the quality in half, and called it an upgrade.
This is straight-up downgrade fraud.
Cancel now. Switch to real models that actually improve.