I'd Rather Just Text My AI

I'd rather just text my AI.

Seriously. I like trying different platforms. Sometimes I want Claude's take, sometimes ChatGPT's, sometimes Gemini's. What I don't particularly enjoy is opening three windows, pasting the same question into each, and then moving the answers around myself.

I have three kids. Repeating myself is already a fairly substantial part of my day.

So here's what I want: one iMessage thread where I can bring in whichever AI I want to talk to.

Claude could answer by default. Type @gpt for an OpenAI model, @gemini for Gemini, or @all when I want all three to weigh in.

Then I could ask:

"@all, which of those ideas is actually worth doing? And what did we miss?"

For that to work, the system would pass their earlier answers into the next round so each model could review what the others said. That's the part I'm curious about. Would one catch a problem the others missed? Would a little pushback help me get somewhere useful?

Yes, I realize I've just described a meeting. For someone who hates meetings, this is a concerning development.

The setup would need a messaging bridge, a small server, and connections to the model APIs. But from my side, I want it to feel like texting. That's it.

Which brings me back to my usual question:

Why is this so complicated, and does it have to be?

I already know how to send a text. Most of the people I'd want to introduce to AI do too. Yet we keep handing them another app, another interface, and another place to start over.

Maybe we should spend a little less time asking which AI people should use, and a little more time asking how they actually want to use it.

If AI is supposed to make our lives easier, why do we keep asking people to change how they work just to use it?