Anyone else satisfied with the current models on offer but just want them to be faster?

I feel like the exhausting context switching work-flow that AI has created could be solved if they were 10x, 100x, 1000x faster. Back to working on one thing at a time please.