After a lot of work restructuring the panel to be able to reliably get a response to a command it runs, I was able to throw together two new commands for the ServerAI.

Time? (or asking "what time it is/is it daytime" etc... )

This will let you know the time of day in game as well as a more easy to read string like "Day" or "Night", great for when you are mining.

TPS? (Ticks Per Second)

This command will take a sample of how fast the server is running compared to what it should be. (20.0tps) It will show the tps along with a color representing how much load the server is under.

Why it took so long

At first glance it looked like an easy thing to do, but as soon as I started, it was very clear it wasn't. It's easy to run a command and forget about the result, but keeping track of the output is much more difficult. You need to know whether or not it succeeded, then get the results back to the place that called on the command without stopping the whole program, all with the knowledge that it could fail at anytime, not respond, give something completely different or even come out of order.

@GodOfDiscordia What is the command system for this?
I've been trying to recreate this on my own world but have had no luck.

I'm mostly stuck at how to use custom phrases to prompt /tellraw click event or general /say command

It's using my own custom panel alongside my own AI chat system.

@GodOfDiscordia Oh my this sounds much more in depth than I originally thought. I'll have to see if i have the muster/knowledge to recreate this.

A basic scoreboard clock system certainly won't do the trick :P

The AI system doesn't use any command blocks except for running in-game announcements like the vote links.

@GodOfDiscordia Hmmm I see, so then in order to implement this you've created your own external code and attached it within the server files/ jar? Or am I still misunderstanding :o

No modification to the jar, there is a program that runs the jar file and watches the output (My panel). This program looks for questions (stuff that starts with who/what/when/where/why or ends with "?" a bit more complicated than that) and sends the questions to my service that resolves them and sends back a response in the form of a command to run.

@Jordi223b If I'm not mistaken, correct me if I'm wrong, it uses a Neural Network, those are used for facial recognition and Go AI too, I think

Yes, it uses a combination of a Neural Network alongside a few other algorithms to find keywords and key statements. It uses that information to find the best match for the question.