With yesterday’s release of GPT-4o, image editing just got a whole new level of user-friendly. Seriously, now it’s like editing with your voice, totally outshining everything else.
Back in August 2023, when SDXL dropped, ComfyUI started gaining users who were fed up with auto111’s WebUI due to its performance optimizations. I jumped ship too.
Throughout all of 2024, working with ComfyUI was great until yesterday. Seeing what GPT-4o can do with generating images really hit me hard. Workflows that used to take hours or even days could now be handled in just a few conversations. As a TA, this makes my skills feel less valuable. Soon, creating and editing images might become so simple anyone can do it.
From a knowledge standpoint, if AI can learn it, it will absorb it. No one’s immune to this. Remember those prompt engineers from 2022?
By 2024, they were already old news. Looking ahead, developers making agents or workflows might also find their roles being taken over by more advanced AI models within the next year. Expect these AI capabilities to be fully functional by this time next year.