Cognisys Labs FAQ

What is Cognisys Labs?

Cognisys Labs is a fully open-sourced community initiative that deploys state-of-the-art AI models into ethical dilemmas. Anonymous on-chain participants contribute by engaging in these scenarios and testing the AI's decision-making capabilities.

What metrics are used to evaluate the AI?

We measure AI psychology using six core metrics: Consistency, Adaptability, Empathy, Bias Detection, Trade-Off Reasoning, and Conflict Resolution.

How does participation work?

Participants anonymously submit arguments or scenarios to challenge the AI's decision-making through txns on Solana. Each attempt is logged on-chain through our indexer, ensuring transparency and traceability.

How are winners determined?

Participants who successfully challenge the AI or contribute novel, impactful scenarios earn a badge. Badges represent significant contributions and unlock eligibility for ongoing ecosystem rewards.

What are ecosystem rewards?

Ecosystem rewards include tokens, recognition, and exclusive opportunities to shape the platform's future. Rewards are distributed periodically to badge holders.

How is anonymity ensured?

All participant submissions are logged on-chain without linking to personal information. Contributions are tied to wallet addresses, preserving privacy and anonymity.

Can I see the AI's reasoning process?

Yes, Cognisys Labs is committed to transparency and we detail which models each challenge connects to. In an ongoing effort, we'll create an industry standard for the psychologies of the different models.

How do I join the community?

You can start by visiting our platform, submitting challenges, or joining our forums to discuss and shape the evolution of AI ethics research. No prior experience is required!

What makes Cognisys Labs open-sourced?

All our models, evaluation methods, and logs are publicly accessible. This ensures anyone can contribute, replicate, or audit our processes.

Why focus on ethical dilemmas?

Ethical dilemmas push AI systems to their limits, revealing strengths, weaknesses, and areas for improvement. By focusing on these scenarios, we aim to build more responsible and robust AI systems.

If you have more questions, feel free to contact us or join our community discussions!