OpenAI’s Real Alignment Problem Has Nothing to Do With AGI

The company understands intelligence at superhuman levels. Human behavior keeps breaking the product.

Continue reading on Towards AI »