An Open Letter to Bernie Sanders: Regulate AI’s Dangers, Don’t Ban Its Promise

Editor's take

The Ban Artificial Superintelligence Act, while acknowledging genuine AI governance shortcomings, proposes a definition that conflates artificial general intelligence (AGI) with superintelligence. This distinction is critical; AGI, the development of which is still theoretical and far from current capabilities like OpenAI's GPT-4 or Google's Gemini, represents a broad spectrum of advanced AI.

The act's broad definition could inadvertently stifle research and development in areas of AI that are already demonstrating societal benefits, from drug discovery to climate modeling. Policymakers must carefully differentiate between the speculative risks of hypothetical superintelligence and the tangible applications of increasingly capable, but still narrowly focused, AI systems.

Future legislative efforts should focus on targeted regulations addressing specific AI risks, such as bias in hiring algorithms or the spread of misinformation, rather than broad prohibitions based on an imprecise understanding of AI's developmental trajectory. The true challenge lies in fostering responsible innovation, not in hindering progress through overly simplistic or misinformed policy.