We're sunsetting PodQuest on 2025-07-28. Thank you for your support!
Export Podcast Subscriptions
cover of episode AI, Agents and Software 3.0
People
主
主持人
专注于电动车和能源领域的播客主持人和内容创作者。
Topics
主持人：我认为软件领域在过去70年里变化不大，但现在在很短的时间内发生了两次重大转变。这体现在工程领域的讨论中，我们现在谈论的AI工程师不仅仅是机器学习研究人员和数据科学家，而是构建在新的生态系统之上，利用基础模型、代理和新工具来重新设计人们与软件交互方式的人。我引用Andrej Karpathy的话，AI工程师的数量可能会大大超过ML工程师。评估、应用和产品化AI面临着无数挑战，这是一项全职工作。软件工程将产生一个新的子学科，专门研究AI的应用，就像站点可靠性工程师、DevOps工程师、数据工程师和分析工程师一样。Context engineering是构建动态系统，以正确的格式提供正确的信息和工具，使LLM能够合理地完成任务。

Deep Dive

Chapters
This chapter introduces Software 3.0, a new era of software development characterized by the use of large language models (LLMs). It explores the evolution of software from Software 1.0 (human-written code) to Software 2.0 (neural networks) and Software 3.0, where LLMs are programmed using natural language. The chapter also discusses the rise of the AI engineer and the challenges of evaluating and applying AI in product development.
  • Software 1.0: Human-written code
  • Software 2.0: Neural networks learned from data
  • Software 3.0: LLMs programmed with natural language prompts
  • Rise of the AI engineer as a new sub-discipline in software engineering
  • Challenges in evaluating, applying, and productizing AI

Shownotes Transcript

Andrej Karpathy's Software 3.0 talk reframes LLMs as a new kind of software—programmable, agent-native, and fundamentally different from past computing models. This episode breaks down his key ideas, from autonomy sliders to the need for new infrastructure designed for AI-first users.

**Source: **https://www.youtube.com/watch?v=LCEmiRjPEtQ)