I am a daily user of GPT-3.5 and for more advanced work, I only use GPT-4. However, today I am noticing subpar performance from GPT-4. I am certain that I am not imagining this. I brainstorm with the system daily and I simply notice that it is not functioning as usual. What is going on?