In the mid-90s, I was one of those muppets who thought that with OS/2 Warp 4 and speech recognition, the future finally arrived. We all know that this was rather disappointing, but well, they tried.
Chat
So we went with chat instead, AIM and ICQ, IRC for techies…Chat, being such a nice hybrid of sync/async communication and written/oral, working on most platforms and also on tiny cell phone screens was the perfect interface those days. And still is for most of us.
ChatOps
ChatOps here was the logical next step, and even the integration of Skype chat bots, CI/CD status messages and uptime data was a logical evolution.
Conversational UIs
Now this year, there’s a new dot on ThoughtWorks’ Technology Radar: Conversational UIs. Go read it, but the TL;DR is that you don’t need a screen and won’t click buttons, but a working speech recognition is all the UI you need.
One logical step would IMHO be a simple pattern recognition, where a chat bot would suggest steps or point to related issues, maybe mention people to involve. But what about actual commands?
Now, let’s combine that with the status quo in ChatOps – what’s next? Thoughts?