This article criticizes several South Dakota legislators for relying heavily on artificial intelligence in their legislative duties, arguing that AI tools are not reliable substitutes for traditional research, constituent interaction, or human judgment, leading to a decline in legislative rigor and potential 'cognitive surrender'.
Senator Novstrup's Defense of AI Use
Senator Al Novstrup defends his use of AI tools like Gemini for legislative work, comparing it to traditional research methods such as using a library or reading newspapers. The author refutes this comparison, arguing that AI merely produces statistically probable text without regard for truth or democratic principles, unlike human sources or direct engagement with constituents.
Other Legislators Embracing AI
The article highlights other legislators, including Rep. Kent Roe, who utilize generative AI tools like Grok for drafting bills and refining ideas. While Roe emphasizes refining queries and cross-checking answers, the author expresses concern about the potential for Musk-biased or reality-detached information from such AI, questioning the reliability of its outputs.
Concerns from Legislative Research Council
John McCullough, director of the Legislative Research Council, expresses strong skepticism about AI's ability to perform trustworthy legal analysis, citing instances where AI 'hallucinated' non-existent laws. He acknowledges AI's role in speeding up legal research tools but stresses its limitations and the irreplaceable need for human oversight to ensure legal consistency and accuracy, especially given South Dakota's small legislative staff.
Erosion of Legislative Quality and 'Cognitive Surrender'
Senator Liz Larson observes a concerning decline in the diligence and rigor of legislative work, directly linking it to the uncritical adoption of AI. The article introduces the concept of 'cognitive surrender,' a phenomenon where individuals accept AI's output at face value, even when incorrect, overriding their own intuition. This suggests that legislators might be forfeiting critical thinking by over-relying on AI.
Call for Human Intellect in Lawmaking
The author concludes by urging elected officials, particularly those seeking re-election, to recognize their fundamental duty to exercise their own intellect, express their own voices, and actively resist the temptation to delegate their critical decision-making responsibilities and identity to artificial intelligence, advocating for a return to human-centered governance.
