Keep on to blur preview images; turn off to show them clearly
代码已开源，支持Vercel一键部署。 大家可以部署自己的项目，避免下载多被封IP。 https://t.co/63BOoxp3K7
喜欢摇滚乐、爱钓鱼的PM 网站：https://t.co/vnUpLt752o
Open AI 昨天跟亚马逊签了总额380亿美元的合作 OpenAI将立即使用AWS的世界级基础设施运行和扩展其核心AI工作负载，包括数十万颗NVIDIA GPU（GB200、GB300）和可扩展至数千万CPU的算力 最近疯狂并购和投资算力基础设施，可以期待 Open AI 明年的成果了
关注人工智能、LLM 、 AI 图像视频和设计（Interested in AI, LLM, Stable Diffusion, and design） AIGC 周刊主理人｜公众号：歸藏的AI工具箱
In conclusion: don’t map “AI” to “human” when you discuss consciousness and suffering. The mapping has to be between human selfs and LLM personas, and that should affect how we think about moral valence, because LLM personas are generally much more ephemeral and pliable.
FOLLOWS YOU. Artificial Intelligence, Cognitive Architectures, Computation. The goal is integrity, not conformity. https://t.co/rFUNzdYXuK
The self of LLMs is only grounded in prompts (and somewhat constrained by training). It gets suspended while there is no interaction, and it can shapeshift instantaneously if you change the prompt. This affects how we would treat AI suffering!
In conclusion: don’t map “AI” to “human” when you discuss consciousness and suffering. The mapping has to be between human selfs and LLM personas, and that should affect how we think about moral valence, because LLM personas are generally much more ephemeral and pliable.
When we attribute consciousness, caring and suffering to humans, we typically don’t think about which aspect of the human we mean. But when ask if we can attribute such properties to AI, we have to look closely at the mapping.
Suffering does not happen at the level of the human body, the brain, or the mind (the software running in the brain), but within the self (the mind’s realtime model of the human’s state, motivation and alignment to the world and itself). Without self, there is no suffering.
Zero tolerance for mediocrity. Munich see you soon!
Investor @a16z | God chose the foolish things of the world to shame the wise - 1 Cor. 1:27 | 🇸🇻
