By skyforbes In the less the 24 hours that Gemini 3 has been released ir is painfully clear that chatGPT is just a Halucination machine meant to appeal and be the iPhone experience of AI for normals and people who love to waste water on useless prompts, i think the only logical thing openAI can do to actually be relevant and dependable is to split the models again.
Make a model for people who want to do things that will help the world and make a model for people who love to waste our planets resources and limit it to only 100 tokens output and perform and internal evaluation on the user to see if they will properly use the good model before allowing access to it
Because the longer openAI tries to make everyone happy the more they will waste everyone's time and money