This is a demo article for bitter-sweet.ai. It explains the core metaphor: AI as something that tastes both sweet and bitter. Sweet, because it can support innovation, medicine, accessibility and creativity. Bitter, because it can deepen inequalities, create new forms of pressure and subtly influence how we see ourselves and others.
MIT economist Daron Acemoglu argues that AI's GDP impact over the next decade will be modest (0.5-1.5%), challenging the hype around transformative economic effects.
Follow-up review documenting significant AI progress in healthcare but also revealing persistent challenges: dataset bias, lack of real-world validation, and regulatory gaps.
A 200+ page report by 100+ Stanford researchers mapping the entire landscape of foundation models — from capabilities to societal risks, bias, and environmental impact.