Gary Marcus分享了他对LLM扩展与AGI关系的最新看法，指出真正进展在于符号化添加而非纯扩展。
Gary Marcus在社交媒体上重申，他认为纯扩展大型语言模型无法实现通用人工智能。他强调，最近的进展实际上不是关于扩展LLM，而是向LLM添加确定性、符号化的内容。Marcus认为这种添加是必要的，但不足以实现AGI。他提到了自己的文章《AI的下一个十年》作为支持观点的依据。
No. I still don't think that pure scaling of LLMs can lead to AGI. BUT... ... most recent advanc...
No. I still don't think that pure scaling of LLMs can lead to AGI. BUT... ... most recent advancements aren't actually about scaling LLMs. The really about adding deterministic, symbolic stuff to LLMs. Even that is not sufficient to get to AGI. But it is necessary. (see my article Next Decade in AI). Egide Murisa @EgideMurisa @GaryMarcus Hi Gary, Given the recent advancements, has your perspective changed regarding whether scaling LLMs can lead to AGI? 🔗 View Quoted Tweet 💬 13 🔄 8 ❤️ 62 👀 10670 📊 17 ⚡