TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Ask HN: Should LLMs be all-knowing experts or rapid-learning minds?

1 点作者 anandnair3 个月前
Should we build LLMs like they&#x27;re all-knowing experts of everything from programming languages to even neuroscience.<p>Is that really the right way to look at them? Or should it be made in such a way that they&#x27;re so good at thinking, planning, and reasoning, with the unique capacity to learn and adapt incredibly fast - much like a fresh human brain?

1 comment

Terr_3 个月前
What makes you think you have a choice? That&#x27;s like asking whether binary heaps should be more ambitious or more polite.<p>Most of what you&#x27;re describing is either faked by something outside the LLM, or if invented later will involve some completely new approach that will have a different name.<p>If you&#x27;re thinking of prompts like &quot;be friendly&quot;, that&#x27;s not imbuing a behavior into the LLM itself, but rather adding a prefix to the document in the hopes it will have a desirable effect on what gets appended.
评论 #42934262 未加载