一位用户回忆起过去在STEM课堂上的经历，当时大型语言模型（LLMs）被呈现为确定性的，尽管它们本质上是概率性的。用户试图强调这些模型相关的巨大计算成本，但同学们却驳回了这一担忧，认为公司不会将如此有用的技术商业化。用户感到欣慰，认为自己预见了当前人工智能的经济现实。 AI
排序理由 用户观点文章，没有新的事实信息。
在 Mastodon — fosstodon.org 阅读 →
AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →
用户回忆早期对大型语言模型成本和确定性演示的怀疑
报道来源 [1]
-
About a year ago i was sitting in a stem 100 class where people presented about LLMs like they were deterministic. I tried to question the presenters in ways th
About a year ago i was sitting in a stem 100 class where people presented about LLMs like they were deterministic. I tried to question the presenters in ways that would point this out but no one understood my point. I tried to bring up the cost of computing as evidenced by data c…