到2030年的AI量化预测集中于工作、医药、电力与行为占比的跃升。调查涵盖300余名专家与公众，专家中位数预测显著高于公众。关键比率包括18%美国工作时数将由生成式AI辅助，以及25%新获批药物销售将来自AI发现药物。7%美国用电将用于AI，约为当前数据中心总耗电的1.5倍。15%成年人将每日使用AI获得陪伴或情感支持，高于当下的6%。

预测分歧形成重要趋势对照。公共预测最显著落后于专家之处在于自动驾驶：专家预测2030年自动驾驶将占美国网约车行程的20%，公众则仅给出12%。专家群体整体比公众更看好AI扩张，但内部仍存在显著意见差异，且整体判断低于部分大型实验室领军人物的激进预期。

总体比率显示AI将在劳动、药物研发、能源需求与行为模式中形成两到数倍级的增长节点，代表结构性社会转变的可量化基础。这些占比投射了2030年前后AI成为多部门关键输入的概率梯度，也为衡量“根本性变革”提供分项指标框架。

AI’s quantified trajectory to 2030 centers on rising shares across labor, pharmaceuticals, electricity, and behavioral adoption. A survey of over 300 experts and members of the public shows that expert medians exceed public expectations. Key ratios include 18% of US work hours assisted by generative AI and 25% of new pharmaceutical sales coming from AI-discovered drugs. AI is projected to consume 7% of US electricity, about 1.5× today’s total data-center use. Daily AI companionship usage is expected to reach 15% of adults, up from 6% today.

Forecast gaps highlight structural divergence. The largest public–expert gap appears in autonomous mobility: experts predict 20% of US rideshare trips will be completed by autonomous vehicles in 2030, while the public predicts 12%. Experts overall anticipate more AI expansion than the public but remain less aggressive than some leading AI-lab figures, with significant intra-expert disagreement.

The ratios collectively indicate two-to-several-fold increases across key domains, forming a measurable basis for structural transformation. These shares outline the probability gradient of AI becoming a critical input across multiple sectors by 2030 and offer a component framework for assessing “fundamental change.”