Part 12/15:
Miikkulainen envisions a future where open AI models support global scientific collaboration, allowing researchers across institutions and countries to analyze, refine, and align AI systems ethically. Initiatives like open consortia and shared compute resources are vital in democratizing access and fostering responsible innovation.
He warns of the risks of closed models and proprietary systems, which can hinder understanding and oversight. Instead, widespread transparency enables better bias detection, value alignment, and safety measures—building trustworthy AI from the foundation up.