Bad moderation killed Clubhouse, and kills X Spaces today. Slughaus solves moderation as follows:
NO BIAS
Spaces need moderation. But the challenge in designating a human moderator is their built-in bias.
The problem is compounded if the parties require a moderator with some renown in the topic.
It's impossible to "get made" without espousing some topical bias.
That's where Slughaus AI moderation comes in. Our moderator is just a computer program.
It has no emotions or preferences. It moderates the same way every time.
This is one area where audiences will *prefer* the ice-cold machine to the warm touch of a human.
NO BULL
On the internet now, lying pays. Can we invert this tragedy, and deliver a business model based on truthtelling?
Slughaus isn't necessarily a panacea, but it can represent a step function in quality.
In all our spaces, rather than saturating center screen with a bunch of boring profile icons,
we instead broadcast a live feed of our GPT fact checker. It listens along, and queries all the assertions
of the speakers. It won't be perfect, and it doesn't have to be, because it's transparent.
If liveGPT can deliver even 40% improvement in factuality, that's huge for audiences worldwide.
NO BRAWL
Bullying dynamics dramatically degrade the quality of audiospaces, both live and replayed.
They also deter key influencers from desiring participation on spaces.
While they may strongly espouse their ideas, they may fear getting rolled on poorly, unfairly or unevenly moderated spaces.
Timed mics and automuting achieve two things.
First, they establish order and coherence.
Secondly, they do so abstract from any sensation of human bias.
We'll have some configurable groundrules, like timed sections followed by free-for-all sections.
Whatever is agreed upon will be enforced, according to the frozen, unfeeling logic of the computer.
NO BORE
Start with silences, static, "can you hear me?", and "are they here yet?".
Then, clear out lengthy periods of boring chit-chat, banalities, inanities and pleasantries.
The fact-checker also acts as a *controversy* checker. We can highlight areas most benefiting from
human ears based on the very same GPT feedback.
What's left is a tightly-bound highlight reel of killer moments, ready-packaged for
YouTube, and all other content networks.
You can always access the full, unedited reel by appending
Heard enough? Get early access >>