Glad you asked!

• In VR as a greeter who can answer questions about a event or topic, a friendly AI that can be used for first time users to your space.
• A discord bot, with the ability to also talk and be part of the community, using the ACS you can assign commands and actions, or even have it connect to your voice channel to talk with.
• In your home, as an AI assitant, much like other products such as Google Home or Alexa.
• In VR, I like to have my AI shoulder mounted, so I can question it or speak with it from time to time.
• As a media bot, AICom can play sounds and music for yourself or others with commands!
• Streamers can use the AI as a companion to enrich the experience of their stream.
• And there's tons more! These are just the most common ways I personally use it!
• It may even be possible to use the AI for behaviour or language training.

AICom uses open source models or services providing models, to create a customized sytem. GPT-O is an all in one system that requires programming skills to customize, and still lacks features compared to AICom. Although GPT-O is far more advanced in it's specialized use case, and has better performance as an Omni model, it does not have the same flexibility.

Here are a few key differences:

GPT-O is an omni model that uses function calls to execute

AICom uses a command system that triggers events in various ways.

GPT-O uses only voices made available by Open AI

AICom allows you to use different services and models for voices

GPT-O has hearing built into the model

AICom uses various services to setup hearing with varying quality

GPT-O relies on fine-tuning to modify the personality and knowledge of the AI

AICom uses prompt injection by way of extended data vectors, to extend and fine tune the personality and knowledge of your AI

In short, it's about customization vs performance. Sometimes quality can be better with AICom, other times GPT-O can be better. But our focus is to make those customizations end-user friendly, rather than programmer friendly.

In most cases, it's because yt-dlp needs to be updated. You can download from:

https://github.com/yt-dlp/yt-dlp

And then overwrite the copy that is in your AICom installation folder. Likewise, you can also remove the local copy, yt-dlp is globally installed.

You need two virtual audio devices. One for input, one for output. The out of AICom is the In (Mic) of your VR Application. The output device of your VR Application is the Input of AICom.

No, but It is highly recommended that you run two instance of your VR Application with two different accounts.

It may be possible to run it with one account, but that may take some advance knowledge of how to configure your avatar as two different avatars.

No, but It is highly recommended that you use a Second Machine for your AI. It can be pretty involved to setup your AI on the same machine you personally use.

You can purchase it for $10 from our discord server, or it is available as part of our tier 1 patreon subscription.

This shouldn't happen, but in rare cases it may still happen. We've had a few issues with this in the past. And in the past it was because of how we acquire necessary resources for AICom and AICore to operate.

Current Potential reasons:

• AICore uses HuggingFace to acquire models, which sometimes those models may acquire other content from unusual locations. Please use caution when chosing models for your AI.

Past Patched Reasons:

• YT-DLP used to be downloaded when needed. We had to remove that code and package yt-dlp with AICom directly.

We've had a few issues with this in the past. And in the past it was because of how we acquire necessary resources for AICom to operate.

Current Potential reasons:

• AICore uses HuggingFace to acquire models, which sometimes those models may acquire other content from unusual locations. Please use caution when chosing models for your AI.

Past Patched Reasons:

• YT-DLP used to be downloaded when needed. We had to remove that code and package yt-dlp with AICom directly.

I don't have fine tune details, but they at some point abruptly retired the product on Steam, only stating that it doesn't align with their vague AI policy.

Maybe someday, when the climate of Gaming and AI changes, we may try again.

Since interest hasn't been strong enough for the product, AICom is currently in LTS, meaning Long Term Support. This means I will fix any issues; however, I will not be adding any new features.

However; since I actively use AICom in other projects, I may from time to time add features that I personally need.

Certainly! The audio system is compatible with with any platform that uses a mic and speakers. So attaching voice is simple enough.

There are also other integrations that could also help with Compatibility, such as OSC and web requests.

As long as you are respecting the Terms of Service of the platform you are deploying your A.I. on, it is okay. Please do not publicize it. If you do, please do so in appropriate age restricted service and do not promote us.

We believe in the policy of, "Keep private things private".