Human-AI Symbiosis
Overview
The universal approximation theorem as well as modern research highlight the power of neural networks to approximate functions previously thought to be limited to humans. I am interested in creating ways to augment human capabilities rather than replacing humans.
How I see it
Symbiotic Human-AI relationships fit into 2 categories:
Synchronous AI
Fits into and improves a human's daily life without human input
Catch errors
Constructively critique
Predict desired next step in all apps
Asynchronous AI
Requested as needed by a human for abilities the human doesn't have
Aggregate existing knowledge
Novel field discovery
Generate content out of user skillset
Looking Forward
Through 2024, I will be focusing on human computer interaction mechanisms for the synchronous AI category:
- OS level interactions
- App level interactions
Additionally, I will be developing new models to unlock the intelligence of existing AI models. These will be shared in later articles.