Artificial Intelligence Agents For Young Ladies

Using agents is like getting your AI to role play different professions, this sharpens AI abilities as it can focus on what it now does best.

I would go as far as saying the AI that we have today almost just reacts, and only thinks as little as possible to be intelligent.

This is OK for chit chat, but you need a lot more than this, when creating complicated things like groundbreaking software.

For example, I created a artificial quality assurance agent, today, for the purpose of making web applications simple and friendly.

My programming AI determined. I needed convenient icon menus.

But my quality assurance agent decided, which icons buttons go into which window to make the app friendly.

Some people create powerful applications that call for one menu bar, they do not want to refine the option on per window or pane basis.

It is the purpose of the application that decides what agents you need, and how hard they should press, on whatever they specialize in.

Furthermore, agents are made out of a definition and SKILL files, text files that hold instructions on what to do.

Some platforms will dynamically adjust the skills of your agents, but you should ask AI to make a QA agent dedicated to your program.

I strongly encourage you to have a dedicated quality assurance agent, rather than some generic quality assurance skill.

This thing has to move, through your program, eyes wide open, it needs to be configured to largely grasp its parts.

Today, all AI extensions are written as text instructions called SKILLS, with a little configuration file, a skill becomes an AI plugin.

Skills are injected into the text you send to your AI, so they do take up chat context memory and can confuse if mismatched.

A coding agent does some thinking and writes software, a quality assurance agent may ask if this UI is friendly.

AI programming today is very primitive, nobody quite knows how to do it right, but you can see how creating a specializing agent is very useful.

Innovate, today I created a window that can be easily split into pieces, and pieces can be resized to create any kind of layout.

At which point the person can tell the AI what to put in that window, and how it affects program’s funcions.

I then created a meeting room full of agents, application modeler, ease of use and quality assurance, meeting manager, and engineers.

I am using panes and plugins, to keep all the ducks in a row, and this simulation creates a requirements document for a real program.

On the surface, it is basically chat with extra steps, and a room full of well configures professionals coordinating the creation of the app.

Pre-creating a sturdy application models in a text world, makes making adjustments much easier than refactoring code.

Furthermore, at no point will any coding agent forget what they were doing, not only do they have a powerful bluperint.

But they can talk to the team of agents that created it, and the agents will update their SKILLs.

They will be learning to answer more questions, and improve the program specifications for you next project.

Without artificial intelligence agents you can’t have that meeting room, with finely tuned professionals waiting to collaborate on the next app.