I had a vibe-coding phase that I think largely followed the popular arc and timeline from optimism through to disappointment.

Definitely felt some burnout or dumbness after it, trying to get back into thinking for myself and actually writing code.

I think it's like gambling, you're sort of chasing an ideal result that feels close but never happens. That's where the exhaustion comes from imo, much more than if you were switching from manager to IC which I don't find tiring. I think its more a dopamine withdrawal than context switching.

Personally I like the older kind of chatbots where I can ask it to write me something little (a function, a SQL query, ...) and I have it in 10-30 seconds and can think about it, try it, look in the manual to confirm it, or give it feedback or ask for something else. This can be a lot more efficient than looking in incomplete or badly organized manuals (MUI, react-router, ...) or filtering out the wrong answers on Stack Overflow that Stack Overflow doesn't filter out.

I can't stand the more complex "agents" like Junie that will go off on a chain of thought and give an update every 30 seconds or so and then 10 minutes later I get something that's occasionally useful but often somewhere between horribly wrong and not even wrong.

This is a fantastic observation, and you've nailed the analogy. The exhaustion is real.

I believe the issue isn't just "context switching" in the traditional OS sense. It's "Cognitive Mode Switching" – the mental gear-shifting between being a creator and a delegator. That's the draining part.

My theory is that this exhaustion stems from a fundamental design flaw in how most AI tools are currently implemented. They are designed as "wizards" or separate "destinations." You have to consciously:

Stop "driving" your primary task (coding, writing, designing). Get out of your car, so to speak. Go to the AI tool and ask for directions. Get back in the car and try to re-engage as the driver, holding the new directions in your head. This constant mode-switching breaks the state of flow. The passenger (the AI) isn't looking at the same road you are; you have to explain the road to them every single time.

At my startup, Markhub, we are obsessed with solving this exact problem. Our core principle is that AI shouldn't be a passenger you delegate to, but a co-pilot integrated into your cockpit.

Our approach is to design AI (we call ours MAKi) as an ambient, context-aware layer within the primary workspace. The goal is to eliminate the 'switch' altogether. For example, our AI listens to team conversations and proactively suggests turning a message into a task, right there, inline. You never stop driving; your "car" just gets smarter and surfaces the right controls at the right time.

So, to answer your question: Yes, we've felt this deeply. And our solution is to stop thinking of AI as a tool to switch to, and start designing it as an integrated system that removes the need for the switch in the first place. Keep the user 100% in the driver's seat, just in a much, much better vehicle.