HN에게 물어보세요: 개방형 LLM이 새로운 오프라인 백과사전인가요?
hackernews
|
|
{'이벤트': '📰', '머신러닝/연구': '📰', '하드웨어/반도체': '📰', '취약점/보안': '📰', '기타 AI': '📰', 'AI 딜': '📰', 'AI 모델': '📰', 'AI 서비스': '📰', 'discount': '📰', 'news': '📰', 'review': '📰', 'tip': '📰'} AI 모델
#ai 모델
요약
There was a time when offline encyclopedias were used heavily as internet was not widely available.
왜 중요한가
본문
There was a time when offline encyclopedias were used heavily as internet was not widely available. I used to install britannica encyclopedia and Microsoft Encarta on my PCs (I had no internet at home at that time) and it had helped me a lot between the years 1999 to 2006.<p>Now we have open-weight LLMs. There are very good models of size < 15GB.</p><p>So I’m wondering:</p><p>- Are open-weight LLMs effectively becoming the next evolution of offline encyclopedias?</p><p>- If so, what replaces the idea of “citation” and “verifiability” in this paradigm, as they do hallucinate time to time ?</p><p>Curious how others here are thinking about this shift, especially folks building offline first educational systems.</p>