Sublime
An inspiration engine for ideas
Every Consulting - AI Strategy & Implementation
every.to
The state of AI in early 2024: Gen AI adoption spikes and starts to generate value
mckinsey.com
转：关于 DeepSeek 的研究和思考 (Archerman Capital)
关于这几天很火的 DeepSeek, 我们 (Archerman Capital) 做了一些研究和思考, 和大家分享, enjoy! 灰色部分是技术细节, 不感兴趣的可略过。
几个事实
1) DeepSeek 不是套壳不是蒸馏美国的大模型。 虽然中国有些大模型是套壳和蒸馏的, 但 DeepSeek 不是。
2) 核心架构还是基于 Transformer, deepseek 在架构、工程设计上进行了创新和工艺提升, 实现效率优化。架构上, 采用了混合专家模型 (MoE)、多头潜注意力 (MLA)、多令牌预测 (MTP)、长链式推理 (CoT)、DualPipe 算法等设计, 并进行了依赖强化学习 (RL)... See more
スマートリング「Oura Ring 4」発表、価格は約5万円で10月15日から出荷 | Forbes JAPAN 公式サイト（フォーブス ジャパン）
David Phelanforbesjapan.com
I’m now sharing (in a follow-up post) links to the tests I conducted comparing Manus AI @ManusAI_HQ to OpenAI’s Deep Research. The prompt was highly complex: evaluating ME/CFS disease, generating new hypotheses, and identifying novel therapeutics, using our recent multi-omics AI paper preprint as the initial context (link also below).
ME/CFS is a... See more
Derya Unutmaz, MDx.com
What's next for AI agentic workflows ft. Andrew Ng of AI Fund
youtube.com
With Sonnet 3.7, Anthropic has clearly placed the focus on coding. Other areas were apparently not particularly important to them.
Incremental improvements in MMLU and GPQA, but a giant leap in SWE. It is clear that Anthropic wants to position Sonnet as the coding AI that it already is.