Related to the other thread on LLMs, I realized that "AI" is misleading. What we've created with LLMs is "AS", Artificial Stupidity.
-
Related to the other thread on LLMs, I realized that "AI" is misleading. What we've created with LLMs is "AS", Artificial Stupidity.
LLMs are perfectly capable of imitating an incompetent human:
- they can make bad plagiarized art like a below-average 11-year-old animation fan
- they can write BS-filled text like a first-day-on-the-job marketing intern
- they can misroute calls and misunderstand bug reports like a burnt-out CSR -
All of this is a damned shame, because the underlying algorithms and inventions are actually really cool, and have a lot of potentially great uses. But those aren't what OpenAI et. al. are doing.
-
Scott Williams 🐧replied to Berkubernetus last edited by
Copyright © 2024 NodeBB | Contributors