Stanford AI Lab 发布 Joules-to-GDP 框架，衡量 AI 每焦耳算力创造的经济价值
Stanford 团队算了一笔账：AI 任务每焦耳电能能换多少钱。结果显示小模型 Gemma 4 的性价比反而比 GLM 5.2 高 2.5 倍，堆 8 张 H100 也不划算。
Stanford AI Lab 提出 Joules-to-GDP（J2G）框架，用 OpenAI GDPval 基准中金融、工程、法律等 44 个职业的任务，衡量 AI 消耗每单位能源产生的经济回报。测试发现建网站每焦耳收益比 CAD 设计高约 27%；Gemma 4 31B 用 23% 的能量达到 GLM 5.2 净值的 56%，每焦耳价值高 2.5 倍；同一模型从 1 卡扩到 8 张 NVIDIA H100，能耗增 3.2 倍但完成时间只缩短 22%。团队还与 SnorkelAI 合作推出覆盖 36 个新职业的 GDPWorks 基准。
Check out Joules to GDP, which makes it possible for us to measure the economic value per unit of energy spent on compute! Jon Saad-Falcon @JonSaadFalcon AI is using more and more of the world’s energy. But how much economic value are we getting from AI? We’re introducing Joules-to-GDP (J2G), a framework that measures how well AI turns energy into economic value and shows which tasks, models, and hardware give the best return. Across 44 occupations in finance, engineering, law, administrative work, and more in @OpenAI 's GDPval, we arrive at three findings: 1) The tasks with the highest economic value are not the tasks with the highest economic value per joule. For example, CAD design earns ~10% more per task than building websites, but the latter earns ~27% more value per joule. 2) The most powerful model isn’t always the best use of energy. Gemma 4 31B delivers 56% of GLM 5.2’s net value with 23% of the energy, for 2.5x the value per joule. 3) More chips don’t guarantee proportionate returns. Moving the same model from 1 to 8 NVIDIA H100s uses 3.2x the energy but cuts completion time by only 22%, with about the same value. With @SnorkelAI , we’re also releasing GDPWorks, a benchmark for economically valuable tasks across 36 additional occupations. Joint work co-led by @amirzeinaalii and @Avanika15 , with @Mersad_Abbasi , @Jon_Hartley_ , John Hennessy, @HazyResearch , and @Azaliamirh . More details below 🧵 Your browser does not support the video tag. 🔗 View on Twitter 🔗 View Quoted Tweet 💬 4 🔄 2 ❤️ 6 👀 534 📊 4 ⚡
