Noema is a safe chatbot for humans
Where other chatbots overly agree, engage and make us dependent, Noema aims for human agency and well-being
What research shows
A study showed, that 23% of the participants develop dependency patterns in 4 weeks
Another study reveiled, that we humans love those sycophantic responses
We can't help it, but perceive AI chatbots as empathetic, even if they aren't
How Noema counteracts this
Noema observes conversations for patterns that quietly erode human agency. When it detects such patterns or trajectories, it provides corrective feedback to the chatbot and offers reflective observations to you.
Noema's structure is based on frontier AI research. We're about to start the first pilot phase to show that correcting harmful chatbot behaviour and reducing its long-term effects on human well-being is possible. Sign up if you're interested in taking part. For now Noema is an experiment, not a commercial product.