We're sunsetting PodQuest on 2025-07-28. Thank you for your support!
Export Podcast Subscriptions
cover of episode AI’s Drawbacks: Environmental Damage, Bad Benchmarks, Outsourcing Thinking — With Emily M. Bender and Alex Hanna

AI’s Drawbacks: Environmental Damage, Bad Benchmarks, Outsourcing Thinking — With Emily M. Bender and Alex Hanna

2025/5/14
logo of podcast Big Technology Podcast

Big Technology Podcast
AI Deep Dive AI Chapters Transcript
People
A
Alex Hanna
批评和分析AI技术，特别是OpenAI的新o1模型和AI炒作指数。
E
Emily M. Bender
Topics
Emily M. Bender: 我认为大型语言模型本质上是一种障眼法，它们利用我们理解语言的能力，让我们误以为其中存在智能。这种障眼法通过聊天机器人的用户界面设计进一步增强，例如使用“我”和“我的”等代词，尽管它们实际上并没有自我意识。更重要的是，这种技术被用于构建各种应用，例如法律助手、医疗诊断系统和个性化辅导，但这些应用可能会取代工人，并掩盖社会保障体系的不足。我认为这种技术存在从底层到顶层的欺骗性。 Alex Hanna: 我认为大型语言模型的能源消耗和环境影响是一个重要问题，但公司在这方面的透明度不足。即使在谷歌内部，对于能源消耗的估算也存在争议，因为他们会声称使用了可再生能源或利用了非高峰时段进行训练，但这些信息并不公开。我认为模型训练耗费大量能源，这已经对现实世界产生了实际影响，例如社区失去水源和电力线路变得不稳定。此外，在孟菲斯，XAI 使用甲烷发电机为一台超级计算机供电，这影响了传统上贫困的黑人社区。我认为这些环境影响被隐藏起来，因为我们通常通过移动设备或电脑访问这些技术，而计算及其环境足迹在云的非物质性中被隐藏起来。

Deep Dive

Chapters
This chapter explores the deceptive nature of large language models, highlighting how they mimic human-like conversation despite lacking genuine understanding. It also touches on the ways this technology is being marketed and used to solve problems it isn't equipped to handle.
  • Large language models are essentially parlor tricks that exploit our capacity for linguistic comprehension.
  • The user interfaces of chatbots enhance this illusion, for instance by using first-person pronouns.
  • These models are being falsely marketed as solutions for various professional fields, from law to medicine, promising worker displacement and social safety net solutions.

Shownotes Transcript

Emily  Bender is a computational linguistics professor at the University of Washington. Alex Hanna is the Director of Research at the Distributed AI Research Institute. Bender and Hanna join Big Technology to discuss what their new book, “The AI‑Con," which they describe as the layered ways today’s language‑model boom obscures environmental costs, labor harms, and shaky science. Tune in to hear a lively back‑and‑forth on whether chatbots are useful tools or polished parlor tricks. We also cover benchmark gaming, data‑center water use, doomerism, and more. Hit play for a candid debate that will leave you smarter about where generative AI really stands — and what comes next.

Enjoying Big Technology Podcast? Please rate us five stars ⭐⭐⭐⭐⭐ in your podcast app of choice.

Want a discount for Big Technology on Substack? Here’s 25% off for the first year, which includes membership to our subscriber Discord: https://www.bigtechnology.com/subscribe?coupon=0843016b)

Questions? Feedback? Write to: [email protected]