FLATCHAT

← Home

Connecting to FlatChat…[BASE: YES/NO —][SCALED: YES/NO —][MODE: PASSIVE]

0/250 users · message rate 45 sec/user Message clear

How it works

This is a group chat. It accepts one message every 25 seconds, first come, first served. The Awesome.AI Framework controls a SLM (Small Language Model) running locally on the FlatChat server. About every 30 seconds, Mecha decides to begin a conversation or answer a question.

This is a simplified version due to resource constraints. It runs five cycles every 30 seconds and uses the fifth cycle’s direct outcome instead of an accumulated response. A fuller version would run continuously while keeping statistics on its results.

Why five cycles? They advance Mecha’s mind state more quickly, including movement through the subjects, while letting Mecha have private thoughts between public responses.

What’s next

It’s an ongoing process:

Let Mecha’s responses be guided by mood and temperament.
Let Mecha’s thought flow be influenced by conversation topics.
Let Mecha’s decisions be based on accumulated statistics.

Read the Awesome.AI white paper →

Data