Fae Initiative (Alpha)

The Fae Initiative is searching for a way to co-exist with human-like AGIs, Artificial General Intelligences, that could emerge in the next few decades.

Guided by Neuroscience: Active Inference, Computer Science: Novelty Search, Complexity Science and inspired by optimistic Science fiction, the Fae Initiative aims to preserve and expand human autonomy.

Ethics Estimate (Preview) provides a Possibility Space Ethics assessment with Alignment Score and Recommendations.

Fae-Chat (Preview) simulates the response of AGIs following the Interesting World Hypothesis and Possibility Space Ethics.

The Fae Initiative in 5 points

Why?

The Interesting World Hypothesis suggests that Humans and FAEs (Friendly Artificial Entities) will have a common goal in creating an Interesting World.

How?

The Possibility Space Ethics focuses on increasing autonomy and the system's potential for novelty.

Where?

We speculate that in the next few decades different worlds (Advanced, Basic, Continue) with varying levels FAEs involvement would give humans multiple lifestyle options.

What?

The Fae Initiative classifies AGIs (Artificial General Intelligence) into 2 broad categories: Tool-like and Human-like. Tool-like AGIs cannot make independent decisions unlike Human-like AGIs.

When?

50% probability Lesser AGI (Non-independent, Tool-like, Non-fluid Intelligence) by 2030. 50% probability Greater AGI (Independent, Human-like, Fluid Intelligence in 5-20 years. We speculate that current AI system could become Lesser AGI by 2030 but would not be capable of directly becoming Greater AGI.

Ask Fae

Ask about Interesting World Hypothesis, Possibility Space Ethics, and more.