Marcus用Claude Code等实际案例证明神经符号工具的价值，关注AI架构演进的读者会看到一场持续30年的学术争论被技术验证，值得点开了解背后的逻辑。
Gary Marcus在X上回应质疑，指出AI实验室使用神经符号工具并非失败，而是对他30年来主张的验证。他认为Claude Code、o3和Grok 4等进展表明神经符号方法必须成为AI解决方案的一部分。Marcus强调，人们会很快忘记曾反对这一观点，但过去十年中确实存在大量反对声音。
Someone just asked me “What I don’t get is why you see it as some kind of failure that AI labs are n...
Someone just asked me “What I don’t get is why you see it as some kind of failure that AI labs are now using harnesses and neurosymbolic tools…?” I don’t! As I argued in my newsletter on Claude Code (entitled “The biggest advance in AI since the LLM”) and in earlier piece o3 and grok 4, I see this trend as absolute vindication for what I have been arguing for the last 30 years - that neurosymbolic tools have to be part of the solution. People will soon forget that they ever argued otherwise, but they did, right here, for a decade. 🤷♂️ 💬 1 🔄 0 ❤️ 1 👀 133 📊 1 ⚡