Getting Started

Welcome to our inaugural post about Argh Studios... we really appreciate you taking the time to visit. We wanted to start out by telling our "origin story", so you can see the ideas that got us started down this path, and so you can better understand our goals and values.

In reading about generative AI & playing around with Large Language Models (LLMs) like ChatGPT, & then watching how we *and others* reacted to the ChatGPT interface, we were struck by a couple of simple observations & ideas:

Faced with an interface they think can respond to anything, a lot of both non-techie and tech-savvy people become "keyboard shy", become concerned about how the thing will interpret what they say and what will come out of the application. People want to be wrapped in a cocoon of safety & trust, and they want to feel an LLM will "Do What I Mean" (also known as "DWIM programming").

The Short-ish Version

Everyone is going for the Golden Ring of General AI, and we aren't there yet... we just want to build very focused niche applications/apps to assist everyday people in accomplishing everyday tasks. The whole thing is summed up in the acronym that is our corporate name: "Assisted Research Generalized Helpers", aka Argh Studios.

Is This "Maybe Real Soon Now" or "Can Do Now"?

Our first idea for a useful, safe, & trustworthy application? How about a storytelling coach and tutor, for grandparents and day care workers and other caretakers of children?

As a proof of concept, we wrote a small stand-alone application in a modern compiled language that tells a machine running a "Large Language Model" to act as that storytelling coach and tutor. Our little application was able to successfully set the context to ensure family-safety, trustworthiness, & DWIM. It allowed us to customize the reading comprehension level, and the cultural tradition, and the level of "spiciness", and so on for the story that we got back.

Like the old saying goes:

"Sometimes it ain't how high the elephant flies, but the fact that it flies at all."

It rapidly became clear to us that our tech-demo could become a real highly scalable service, and thus get Argh Studios started up, so we have begun "building out" a multi-user server from scratch. We hope to have a "minimal viable product" ready for testing by everyday folks by the end of November 2024.

We think there are a zillion other such very focused niche applications that the existing technology can already make work "good enough" today... not next year, today... and assist everyday folks in doing everyday things. Each Argh Studio will focus on *one* audience and *one* need.

Signup Call for Early Testers

Are you interested in becoming an early tester of Argh Studios Storyteller? Please contact us at the following address to go on the list, and we will get back to you when we move to the next stage of testing:

storyteller@arghstudios.com

And welcome again, it should be a fun ride...

The Argh Studios Team