#GPT-4o - dumber model

1 messages · Page 1 of 1 (latest)

maiden windBOT
#

Reported by @azure ledge

Bug Report: GPT-4o - dumber model
`Steps to Reproduce`

  • use legacy model in browser or app

  • send a prompt

  • realize that quality is not the same and answers are shorter and worse

`Expected Result`

Same GPT-4o as it was before release of GPT-5. Same quality, same knowledge, same skills, same length of answers. Please bring back the actual GPT-4o and not just GPT-4o mini as it seems what it is.

`Actual Result`

Dumber and shorter answers, not remembering context of conversion, maybe different, dumber version of GPT-4o.

`Environment`

Browser, Android App

#
Additional Information

Please provide relevant details to help resolve the issue, such as:

  • ChatGPT Shared Link (if applicable).
  • Screenshots or videos demonstrating the problem.

-# ➜ Need to contact support? Visit the OpenAI Help Center.

tropic sentinel
#

I noticed this too... I thought it was just me :/

azure ledge
#

I guess that OpenAI just lost the way. Let's see if they will find it again.

azure ledge
#

OpenAI should officially out themselves as criminals, it's just not the real GPT-4o that they brought back, even though they promised, it's just a name in the UI, but what model? Nobody knows.
=> In parallel, no help from "support" that is not up to date, accordingly no feedback is accepted
They really do everything to fall behind at the market, with acting unimaginably dumb, and to die as a company in the end.
Maybe they're just for real ill and don't know what they're actually doing. In any case, their actions don't seem well thought and controlled.